Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Application Priority Framework for Fixed Mobile Converged Communication
Networks
A PhD Thesis Submitted to: Electronic and Computer Engineering Division
School of Engineering and Design Brunel University West London, UK
By Saqib Rasool Chaudhry
Supervised by: Professor Hamed S. Al-Raweshidy
2010/2011
II
Abstract
The current prospects in wired and wireless access networks, it is becoming increasingly important to
address potential convergence in order to offer integrated broadband services. These systems will need
to offer higher data transmission capacities and long battery life, which is the catalyst for an ever-
increasing variety of air interface technologies targeting local area to wide area connectivity.
Current integrated industrial networks do not offer application aware context delivery and enhanced
services for optimised networks. Application aware services provide value-added functionality to
business applications by capturing, integrating, and consolidating intelligence about users and their
endpoint devices from various points in the network. This thesis mainly intends to resolve the issues
related to ubiquitous application aware service, fair allocation of radio access, reduced energy
consumption and improved capacity. A technique that measures and evaluates the data rate demand to
reduce application response time and queuing delay for multi radio interfaces is proposed. The
technique overcomes the challenges of network integration, requiring no user intervention, saving
battery life and selecting the radio access connection for the application requested by the end user.
This study is split in two parts. The first contribution identifies some constraints of the services
towards the application layer in terms of e.g. data rate and signal strength. The objectives are achieved
by application controlled handover (ACH) mechanism in order to maintain acceptable data rate for
real-time application services. It also looks into the impact of the radio link on the application and
identifies elements and parameters like wireless link quality and handover that will influence the
application type. It also identifies some enhanced traditional mechanisms such as distance controlled
multihop and mesh topology required in order to support energy efficient multimedia applications. The
second contribution unfolds an intelligent application priority assignment mechanism (IAPAM) for
medical applications using wireless sensor networks. IAPAM proposes and evaluates a technique based
on prioritising multiple virtual queues for the critical nature of medical data to improve instant
transmission. Various mobility patterns (directed, controlled and random waypoint) has been
investigated and compared by simulating IAPAM enabled mobile BWSN. The following topics have
been studied, modelled, simulated and discussed in this thesis:
1. Application Controlled Handover (ACH) for multi radios over fibre
2. Power Controlled Scheme for mesh multi radios over fibre using ACH
3. IAPAM for Biomedical Wireless Sensor Networks (BWSN) and impact of mobility over IAPAM
enabled BWSN
Extensive simulation studies are performed to analyze and to evaluate the proposed techniques.
Simulation results demonstrate significant improvements in multi radios over fibre performance in
terms of application response delay and power consumption by upto 75% and 15 % respectively,
reduction in traffic loss by upto 53% and reduction in delay for real time application by more than 25%
in some cases.
III
Acknowledgment
First and foremost, I would like to thank the Creator and Ahl al-Bayt, without the blessings and mercy of whom I would not have made it. I would like to commend the efforts of my supervisor Prof. Hamed S Al-Raweshidy who has not only dealt with me professionally but also has been like my brother and parental figure. I shall never forget the level of maturity and professionalism, which he showed during my stay at Brunel University. I am grateful to my parents, Dr. Ghulam Rasool Chaudhry and Mrs. Naseem Akhtar, for all their prayers and for giving me the gifts of life and education. I couldn’t have reached what I have achieved in life without their sacrifices. I can never forget they instilled in me values of conviction, persistence, and hard work. They have always been a source of inspiration and courage for me and I can never thank them or repay them enough for the virtues they have given me. Finally, I would like to dedicate this thesis to everyone I mentioned above and all those who prayed for me throughout my research and have been supporting, encouraging, sacrificing for me especially my sisters and better half.
IV
Table of Contents Chapter 1: Introduction 1 1.1 WIRELESS COMMUNICATION SYSTEMS ......................................................... 1 1.2 BROADBAND WIRELESS COMMUNICATION SYSTEM ................................. 3 1.3 CHALLENGES OF BROADBAND WIRELESS ACCESS NETWORK ............... 4 1.4 RADIO OVER FIBRE TECHNOLOGY .................................................................. 7 1.4.1. WHAT IS ROF .......................................................................................................... 7 1.4.2. ADVANTAGES OF ROF ......................................................................................... 9 1.4.3. LIMITATIONS OF ROF......................................................................................... 13 1.5 INTEGRATING WIRELESS AND FIBRE OPTICS ............................................. 13 1.6 THE RESEARCH AIM AND CONTRIBUTIONS ................................................ 16 1.7 APPLICATIONS AND THE FUTURE NETWORKS........................................... 18 1.8 OBJECTIVES AND RESEARCH METHODOLOGY .......................................... 19 1.8.1. SIMULATION ENVIRONMENT .......................................................................... 20 1.8.2. MODEL DESIGN.................................................................................................... 20 1.8.3. COMPARISON OF SIMULATION RESULTS..................................................... 20 1.9 THE THESIS STRUCTURE................................................................................... 21 1.10 REFERENCES ........................................................................................................ 22 Chapter 2: Background and Literature Review 27 2.1 RADIO OVER FIBRE TECHNOLOGIES ............................................................. 27 2.1.1. OPTICAL TRANSMISSION LINK........................................................................ 28 2.2 WIRELESS MOBILE TECHNOLOGIES .............................................................. 29 2.2.1. THE IEEE 802.X WIRELESS NETWORKS ......................................................... 29 A. IEEE 802.11 STANDARDS (WIFI)........................................................................ 29 2.2.2. IEEE 802.15............................................................................................................. 32 A. IEEE 802.15.3.......................................................................................................... 34 B. IEEE 802.15.4.......................................................................................................... 34 2.2.3. ULTRA WIDEBAND (UWB) ................................................................................ 36 2.2.4. IEEE 802.16............................................................................................................. 36 2.3 WIRELESS NETWORKING CHARACTERISTIC COMPARISON ................... 37 2.3.1. WIFI VS. BLUETOOTH......................................................................................... 38 2.3.2. WIFI VS. 3G/UMTS................................................................................................ 39 2.3.3. WIMAX VS. 3G/UMTS.......................................................................................... 40 2.3.4. WIFI VS. WIMAX .................................................................................................. 40 2.3.5. WIRELESS CONVERGENCE INFRA .................................................................. 41 2.4 HANDOVER IN WIRELESS MOBILE NETWORKS.......................................... 42 2.4.1. INTRODUCTION TO HANDOVER...................................................................... 43 2.4.2. HANDOVER METHODOLOGY ........................................................................... 46 A. HANDOVER STRATEGIES .................................................................................. 46 B. HANDOVER ALGORITHMS................................................................................ 46 2.4.3. HANDOVER ISSUES............................................................................................. 47 A. ARCHITECTURAL ISSUES.................................................................................. 48 B. HANDOVER DECISION ALGORITHMS ............................................................ 49 C. HANDOVER-RELATED RESOURCE MANAGEMENT.................................... 50 2.5 CONCLUSIONS...................................................................................................... 53 2.6 REFERENCES ........................................................................................................ 53 Chapter 3: Application Controlled Handover Architecture 59 3.1 INTRODUCTION ................................................................................................... 59 3.2 GLOBAL CONTRIBUTION AND RELATED WORK ........................................ 62 3.3 CURRENT WWIN CAPABILITIES ...................................................................... 63
V
3.3.1. WIRED WIRELESS CONVERGENCE ERA ........................................................ 63 3.3.2. THE RESEARCH CHALLENGE........................................................................... 64 3.4 PROPOSED MODIFICATION IN WWIN ARCHITECTURE ............................. 65 3.4.1. CONSTITUTION OF COMMUNICATION ENVIRONMENT FOR FMC.......... 65 3.4.2. APPLICATION CONTROLLED HANDOVER CLASSIFICATION................... 66 3.5 INFORMATION ELEMENT OF THE PROPOSED ACH .................................... 67 3.5.1. APPLICATION CONTROLLED NETWORK PROTOCOL STACK .................. 67 3.5.2. SELECTION OF RANDOM APPLICATION BY PROBABILITY FUNCTION 68 3.5.3. BATTERY LIFETIME MODEL OF THE WIRELESS DEVICE.......................... 70 3.5.4. BATTERY LIFETIME SIMULATION RESULTS................................................ 72 3.5.5. APPLICATION CONTROLLED HANDOVER NETWORK COMPONENTS ... 74 3.5.6. APPLICATION CONTROLLED HANDOVER PROCESS MODEL................... 76 3.6 PROPOSED NETWORK MODEL LAYOUT ....................................................... 77 3.6.1. MULTIPLE RADIOS OVER FIBRE NETWORK................................................. 78 3.6.2. MULTIPLE RADIOS OVER FIBRE NETWORK WITH ACH............................ 79 3.6.3. NETWORK PARAMETERS .................................................................................. 79 3.7 PERFORMANCE EVALUATION AND DISCUSSIONS..................................... 80 3.7.1. NETWORK THROUGHPUT ................................................................................. 81 3.7.2. HTTP CLIENT SERVER RESPONSE TIME ........................................................ 81 3.7.3. DATABASE CLIENT SERVER RESPONSE TIME............................................. 82 3.7.4. VIDEO CONFERENCING RESPONSE TIME ..................................................... 83 3.7.5. NETWORK MULTIPLE MEDIA ACCESS DELAY ............................................ 84 3.7.6. NETWORK PACKET DROP ................................................................................. 84 3.7.7. NETWORK BER AND POWER CONSUMPTION .............................................. 85 3.8 CONCLUSIONS...................................................................................................... 87 3.9 REFERENCES ........................................................................................................ 87 Chapter 4: Power Controlled Meshed Multi Radios over Fibre 91 4.1 INTRODUCTION ................................................................................................... 91 A. RESEARCH CHALLENGE.................................................................................... 93 4.2 GLOBAL CONTRIBUTION AND RELATED WORK ........................................ 94 4.3 PROPOSED NETWORK FRAMEWORK ............................................................. 95 4.4 DEVELOPMENT OF CORE NETWORK COMPONENTS ................................. 96 A. CROSS LAYER NETWORK PROTOCOL STACK ............................................. 97 B. MULTI RADIO NETWORK CONNECTION SCHEMATIC ............................... 98 C. MULTI RADIOS OVER FIBRE UPLINK/DOWNLINK DESIGN ...................... 99 D. MESHED NETWORK ARCHITECTURE........................................................... 100 E. DISTANCE CONTROLLED POWER EFFICIENCY ......................................... 102 I. Power Saving using Multiple Hops ....................................................................... 103 II. Transmission and receiving power ........................................................................ 104 4.5 NETWORK SETUP AND LAYOUT ................................................................... 106 A. WIRELESS BASE MODEL.................................................................................. 107 B. MULTIPLE RADIOS OVER FIBRE MODEL USING ACH.............................. 107 C. POWER CONTROLLED MESHED MULTIPLE RADIOS OVER FIBRE MODEL107 4.6 PERFORMANCE EVALUATION AND DISCUSSIONS................................... 109 4.6.1. NETWORK THROUGHPUT ............................................................................... 110 4.6.2. END-TO-END NETWORK DELAY.................................................................... 111 4.6.3. NETWORK MULTIPLE MEDIA ACCESS DELAY .......................................... 111 4.6.4. PACKET DROP .................................................................................................... 112 4.6.5 HTTP RESPONSE TIME...................................................................................... 113 4.6.6. FTP RESPONSE TIME......................................................................................... 113 4.6.7. VIDEO CONFERENCING RESPONSE TIME ................................................... 114 4.6.8. END-TO-END POWER CONSUMPTION .......................................................... 115 4.7 CONCLUSIONS.................................................................................................... 115
VI
4.8 REFERENCES ...................................................................................................... 116 Chapter 5: Application Priority Assignment for BWSN 119 5.1 INTRODUCTION ................................................................................................. 119 5.2 GLOBAL CONTRIBUTION AND RELATED WORK ...................................... 121 5.3 CHALLENGES AND PROBLEM STATEMENT ............................................... 124 5.3.1. PROBLEM STATEMENT.................................................................................... 125 5.3.2. THE METHODOLOGY........................................................................................ 125 5.3.3. CROSS LAYER DESIGN FOR IAPAM .............................................................. 126 5.4 INTELLIGENT APPLICATION PRIORITY ASSIGNMENT MECHANISM... 127 5.5 FUNCTIONAL COMPONENTS OF APPLICATION PRIORITY ASSIGNMENT128 5.5.1. PACKET HEADER EXTENSION ....................................................................... 128 5.5.2. MODIFICATION TO CLASSIFIER .................................................................... 134 5.5.3. INTERFACE QUEUE STRUCTURE................................................................... 135 5.5.4. QUEUING DELAY............................................................................................... 136 5.6 PERFORMANCE EVALUATION AND DISCUSSIONS................................... 139 5.6.1. NETWORK THROUGHPUT ............................................................................... 140 5.6.2. NETWORK QUEUING DELAY.......................................................................... 140 5.6.3. NETWORK QUEUING DELAY FOR MULTIPLE APPLICATIONS............... 141 5.6.4. NETWORK DATA LOSS..................................................................................... 144 5.6.5. NETWORK END-TO-END DELAY.................................................................... 144 5.6.6. GEOMETRIC MEAN OF THE QUEUING DELAY........................................... 145 5.6.7. ECG & TEMP QUEUING DELAYS COMPARISON......................................... 146 5.7 CONCLUSIONS.................................................................................................... 147 5.8 REFERENCES ...................................................................................................... 147 Chapter 6: Mobility in IAPAM enabled BWSN 152 6.1 INTRODUCTION ................................................................................................. 152 6.2 COMPONENTS OF BWSN.................................................................................. 154 6.2.1. COMPONENTS OF SENSOR NODE (HARDWARE) ....................................... 154 6.2.2. COMPONENTS OF SENSOR NODE (SOFTWARE)......................................... 154 6.2.3. THE MEDIA ACCESS LAYER (MAC) .............................................................. 155 6.3 GLOBAL CONTRIBUTION AND RELATED WORK ...................................... 158 6.4 PROPOSED NETWORK SCENARIOS FOR BWSN USING IAPAM .............. 163 6.4.1. NETWORK MODEL MOBILITY STRATEGIES............................................... 164 A. Random Mobility ................................................................................................... 164 B. Directed Mobility (predictable) ............................................................................. 165 C. Controlled Mobility ............................................................................................... 165 6.4.2. MOBILITY CHALLENGES IN PROPOSED BWSN.......................................... 165 6.4.3. RANDOM MOBILITY MODEL FOR PROPOSED BWSN ............................... 166 6.4.4. GTS AND CAP TRAFFIC SUPPORT IN BWSN (BEACON ENABLED MODE)167 6.5 SIMULATION ENVIRONMENT ........................................................................ 167 6.6 PERFORMANCE EVALUATION AND DISCUSSIONS................................... 168 6.6.1. AVERAGE NETWORK THROUGHPUT ........................................................... 169 6.6.2. AVERAGE NETWORK DELAY......................................................................... 170 6.6.3. AVERAGE POWER RECEIVED......................................................................... 172 6.7 CONCLUSIONS.................................................................................................... 173 6.8 REFERENCES ...................................................................................................... 174 Chapter 7: Conclusions and Future Work 178 7.1 MAIN CHALLENGES AND SOLUTIONS......................................................... 178 7.2 FUTURE WORK................................................................................................... 180 PUBLICATIONS ............................................................................................................ 182
VII
PUBLISHED: ........................................................................................................ 182 ACCEPTED:.......................................................................................................... 183 SUBMITTED:........................................................................................................ 183 BOOK CHAPTERS:.............................................................................................. 183
VIII
List of Figures Figure 1.1: Growth of Fixed and Mobile Communications. _____________________________ 1 Figure 1.2: Current and Future Wireless Communication Systems. _______________________ 3 Figure 1.3: Components of Narrowband Wireless Access Network. ______________________ 5 Figure 1.4: The RoF Concept. ____________________________________________________ 8 Figure 1.5: 900MHz Fibre-Radio System. ___________________________________________ 9 Figure 2.1: Radio over Fibre System. _____________________________________________ 28 Figure 2.2: Optical Transmission Link. ____________________________________________ 29 Figure 2.3: IEEE 802 in the OSI Layers. ___________________________________________ 30 Figure 2.4: Handover Scenarios. _________________________________________________ 44 Figure 2.5: Handover Issues in the Multiple Wireless Networks. ________________________ 45 Figure 2.6: Controlled Handover Algorithm.________________________________________ 47 Figure 2.7: Architectural Issues. _________________________________________________ 50 Figure 2.8: Hard Handover. _____________________________________________________ 50 Figure 2.9: Decision Algorithms in Handover. ______________________________________ 51 Figure 2.10: Handover Resource Management.______________________________________ 52 Figure 3.1: An Integrated Wired Wireless Network Environment. _______________________ 61 Figure 3.2: Hierarchical Classification of Wired Wireless Network Architectures. __________ 62 Figure 3.3: Proposed FMC Architecture for Wired Wireless Integrated Network. ___________ 65 Figure 3.4: Application Controlled Network Protocol Stack. ___________________________ 68 Figure 3.5: Battery Power Management Flowchart. __________________________________ 69 Figure 3.6: Charger Model Flowchart._____________________________________________ 70 Figure 3.7: Battery Power in Mobile Devices._______________________________________ 73 Figure 3.8: The Network Components of ACH. _____________________________________ 75 Figure 3.9: Data and Socket Control Components of ACH. ____________________________ 76 Figure 3.10: ACH Process Model Flow. ___________________________________________ 77 Figure 3.11: Network Simulation Model Layout. ____________________________________ 78 Figure 3.12 Network Average Throughput. _________________________________________ 81 Figure 3.13: HTTP Average Response Time. _______________________________________ 82 Figure 3.14: FTP Average Response Time. _________________________________________ 83 Figure 3.15: Video Conferencing Average Response Time. ____________________________ 83 Figure 3.16: Network Average Media Access Delay. _________________________________ 84 Figure 3.17: Network Average Packet Drop.________________________________________ 85 Figure 3.18: Network Average BER. ______________________________________________ 86 Figure 3.19: Network Average Power Consumption. _________________________________ 86 Figure 4.1: Proposed Power Controlled Mesh Multi Radio Network._____________________ 93 Figure 4.2: Classification of Wired Wireless Converged Network Architectures. ___________ 95 Figure 4.3: Proposed Multi Radio Network Protocol Stack. ____________________________ 97 Figure 4.4: Multi Radio Network Connection Control. ________________________________ 98 Figure 4.5 a: Multi Radios Design. _______________________________________________ 99 Figure 4.5 c: Multi Radios Uplink. ______________________________________________ 100 Figure 4.6: Mesh Network Architecture. __________________________________________ 100 Figure 4.7b: Distance based Power Controlled Schema. ______________________________ 103 Figure 4.8: Power Controlled Mesh Work-Flow. ___________________________________ 108 Figure 4.9: Network Average Throughput. ________________________________________ 110 Figure 4.10: Network Average End-to-End Delay. __________________________________ 111 Figure 4.11: Network Average Medium Access Delay. ______________________________ 112 Figure 4.12: Network Average Packet Drop._______________________________________ 112 Figure 4.13: HTTP Average Response Time. ______________________________________ 113 Figure 4.14: FTP Average Response Time. ________________________________________ 114 Figure 4.15: Video Conferencing Average Response Time. ___________________________ 114 Figure 4.16: Network Average Power Consumption. ________________________________ 115
IX
Figure 5.1: Classification of Health Monitoring Applications. _________________________ 123 Figure 5.2: Cross Layer Design for IAPAM. _______________________________________ 126 Figure 5.3: IAPAM Network Scenario. __________________________________________ 128 Figure 5.4: Structure of Target Packet. ___________________________________________ 129 Figure 5.5: Structure of Sensor Packet.___________________________________________ 129 Figure 5.6: Priority Assignment Flowchart.________________________________________ 131 Figure 5.7: Application Priority Assignment Functional Block. _______________________ 133 Figure 5.8: Classifier Algorithm Design. __________________________________________ 134 Figure 5.9: Virtual Queue Interface for IAPAM.____________________________________ 135 Figure 5.10: Network Simulation Scenario.________________________________________ 139 Figure 5.11: Network Average Throughput Comparison. _____________________________ 140 Figure 5.12: Network Average Queuing Delay. ____________________________________ 141 Figure 5.13: Comparison of Queuing Delay of Multiple Applications.___________________ 142 Figure 5.14: Comparison of ECG Queuing Delay. __________________________________ 142 Figure 5.15: Comparison of BP Queuing Delay. ____________________________________ 143 Figure 5.16: Comparison of Temperature Queuing Delay. ____________________________ 143 Figure 5.17: Network Average Data Loss (Bytes). __________________________________ 144 Figure 5.18: Network Average End-to-End Delay. __________________________________ 145 Figure 5.19: Geometric Mean of the Queuing Delay at Intermediate Nodes. ______________ 145 Figure 5.20: Comparisons of ECG & TEMP Queuing Delays. _________________________ 146 Figure 6.1: BWSN Proposed Architecture. ________________________________________ 153 Figure 6.2: Biomedical Sensor Node. ____________________________________________ 154 Figure 6.3: Network Layers of IEEE 802.15.4. _____________________________________ 155 Figure 6.4: IEEE 802.15.4 MAC Operational Modes ________________________________ 157 Figure 6.5: Average Network Throughput (10 Nodes). _______________________________ 169 Figure 6.6: Average Network Throughput (50 Nodes). _______________________________ 170 Figure 6.7: Average Network Delay (10 Nodes). ___________________________________ 171 Figure 6.8: Average Network Delay (50 Nodes). ___________________________________ 171 Figure 6.9: Average Rx Power (10 Nodes). ________________________________________ 172 Figure 6.10: Average Rx Power (50 Nodes). _______________________________________ 173
X
List of Tables Table 1.1: WLAN Standards _____________________________________________________ 2 Table 1.2: Frequencies for Wireless Broadband Communications ________________________ 7 Table 2.1: 802.11 Wireless Network Standards [3] ___________________________________ 31 Table 2.2: IEEE 802.15.3 vs. WLAN and Bluetooth [12] ______________________________ 33 Table 2.3: Wireless Technologies Comparison [23] __________________________________ 38 Table 3.1: Classification of Handover _____________________________________________ 66 Table 3.2: Power Consumption Rates in Battery-operated Devices ______________________ 72 Table 3.3: Network Simulation Parameters _________________________________________ 79 Table 3.4: Network Data Traffic Profiles___________________________________________ 80 Table 4.1: Network Application Traffic Profiles ____________________________________ 109 Table 4.2: Network Simulation Environment ______________________________________ 109 Table 6.1: The Simulation Parameters ____________________________________________ 168
XI
Abbreviations and Acronyms
ABC Always Best Connected ACH Application Controlled Handover AP Access Point BAN Body Area Network BB Baseband BER Bit Error Rate BI Beacon Interval BP Blood Pressure BS Base Station BSS Basic Service Set BWA Broadband Wireless Access BWSN Biomedical Wireless Sensor Network CDMA Code Division Multiple Access CFP Contention Free Period CO Central Office CS Central Station CSMA/CA Carrier Scenes Multiple Access with Collision Avoidance DAC Digital to Analogue Convertor DAS Distributed Antenna System DCF Distributed Coordination Function DECT Digital Enhanced Cordless Telecommunications DFB Distributed Feedback DR Dynamic Range DSP Digital Signal Processing DWDM Dense Wavelength Division Multiplex ECG Electrocardiogram EDFA Erbium Doped Fibre Amplifier EDR Enhanced Data Rate EEG Electroencephalograph EMG Electromyogram EMI ElectroMagnetic Interference EOG Electrooculography FBG Fibre Bragg Gratings FCC Federal Communications Commission FMC Fixed Mobile Convergence FTP File Transfer Protocol FWA Fixed Wireless Access GPRS General Packet Radio Services GSM Global System for Mobile Communication GTS Guaranteed Time Slot HFR Hybrid Fibre Radio HR High Rate HSDPA High-Speed Downlink Packet Access HTTP Hyper Transmission Transfer Protocol IA Intelligent Algorithm IAPAM Intelligent Application Priority Assignment Mechanism
XII
IBSS Independent Basic Service Set ICT Information and Communication Technologies IETF Internet Engineering Task Force IMDD Intensity Modulation and Direct Detection IP Internet Protocol ISM The industrial, Scientific and Medical ISP Internet Service Provider JSIM Java-based simulation system LAN Local Area Network MAC Media Access Layer MANET Mobile Ad hoc Network MAS Medium Access Slot Mbps Mega Bits Per Second MBWA Metropolitan Broadband Wireless Access MCHO Mobile-Controlled Handover MDEV Mesh Device MMF Multi Mode Fibre MNC Mesh Network Coordinator MPNC Mesh Pico Net Coordinator MU Mobile Unit MUX Multiplexer NCHO Network Controlled Handover NF Noise Figure NIC Network Interface Card NS2 Network Simulator 2 OFDM Orthogonal Frequency-Division Multiplexing OLT Optical Line Terminal ONU Optical Network Unit OS Operating System OSI Open System Interconnection. OTDM Optical Time Division Multiplexing PCB Probability of Blocking Connections PDA Personal Digital Assistant PDF Probability Distribution Function PER Packet Error Rate PHD Probability of Dropping Handovers PHY Physical Layer PNC Piconet Coordinator POF Polymer Optical Fibre PON Passive Optical Network Pos Position PSRC Power Source QAM Quadrature Amplitude Modulation QoS Quality of Service RAM Random Access Memory RAP Radio Access Point RAU Radio Access Unit RF Radio Frequency RFD Reduced-Function Device RoF Radio over Fibre
XIII
Rx Receiver SCM Sub-Carrier Multiplexing SD Service Discovery SDR Software Defined Radio SEC Security SIG Special Interest Group SIM Subscriber Identity Module SMF Single Mode Fibre TCP Transmission Control Protocol TDM Time Division Multiplexing TDMA Time Division Multiple Access TRANS Transport Layer Tx Transmitter UMTS Universal Mobile Telecommunication System UWB Ultra Wideband VPN Virtual Private Network WDM Wavelength Division Multiplexing WiMAX Worldwide Interoperability for Microwave Access WLAN Wireless Local Area Network WMAN Wireless Metropolitan Area Network WPAN Wireless Personal Area Network WTU Wireless Terminal Unit WWIN Wired Wireless Integration Network
1
Chapter 1
Introduction __________________________________________________________
1.1 Wireless Communication Systems
Wireless communication has experienced tremendous growth in the last
decade. In 1991 less than 1% of the world’s population had access to a mobile phone.
By the end of 2001, an estimated one in every six people had a mobile phone [1].
During the same period the number of countries worldwide having a mobile network
increased from just three to over 90%. In fact the number of mobile subscribers
overtook the number of fixed-line subscribers in 2002. It is predicted that this growth
will continue to rise, and by 2016 there will be more than 8 billion mobile subscribers
worldwide [2] as shown in Figure 1.1.
Figure 1.1: Growth of Fixed and Mobile Communications [2].
Apart from mobile telephone communications, Wireless Local Area Networks
(WLANs), which came on the scene less than a decade ago (1997), have also
experienced phenomenal growth. The rapid proliferation of WLAN hotspots in public
2
places, such as airport terminals has been astounding. In fact WLANs have now made
their way into homes, riding on the back of xDSL and cable access modems, which
are now integrated with WLAN Radio Access Points (RAPs). As a result, the number
of wireless Internet subscribers is expected to overtake the number of wired Internet
users quite soon, as shown in Figure 1.1. The growth of wireless data systems is also
seen in the many new standards, which have recently been developed or are currently
under development as discussed in Section 1.2.
The rapid growth of wireless communications is mainly attributed to their ease
of installation in comparison to fixed networks [1]. However, technological
advancement, and competition among mobile operators have also contributed to the
growth. So far there have been three mobile telephone standards, launched in
succession approximately every decade. The first-generation (1G) mobile systems
were analogue, and were commissioned in the 1980s. In the 1990s, second-generation
(2G) digital mobile systems such as the Global System for Mobile communications
(GSM) came on the scene. The GSM standard has been extremely successful,
providing not only national, but also international coverage as well. Universal Mobile
Telecommunication System (UMTS) is currently the mainstream mobile
communication system referred as third-generation (3G).
Table 1.1: WLAN Standards Year WLAN
Standard
Frequency Modulation Bit Rare
(Max) 1997 IEEE 802.11 2.4 GHz Frequency Hopping and Direct
Spread Spectrum
2 Mbps
1998 Home RF 2.4 GHz Wideband Frequency Hopping 1.6 Mbps
1999 IEEE 802.11b 2.4 GHz Direct Sequence Spread Spectrum 11 Mbps
1999 IEEE 802.11a 5GHz OFDM 54 Mbps
2000 HiperLAN 2 5 GHz OFDM Connectionoriented 54 Mbps
2003 IEEE 802.11g 2.4 GHz OFDM compatible with 802.11a 54 Mbps
2007 IEEE 802.11n 2.4/5 GHz Quadrature Amplitude Modulation 600 Mbps
3
Both 1G and 2G systems were designed primarily to provide voice
applications, and to support circuit-switched services [4]. However, GSM does offer
data communication services to users, although the data rates are limited to just a few
tens of kbps. In contrast, WLANs originally designed to provide fixed data network
extension, support Mbps data transmission rates. The WLAN standard – IEEE 802.11,
also known as Wi-Fi, was first commissioned in 1997 and offered 2 Mbps. Since then,
the standard has evolved several times responding to the sustained user demand for
higher bit-rates as shown in Table 1.1. Currently, WLANs are capable of offering up-
to 54 Mbps for the IEEE 802.11a/g, and HiperLAN2 standards operating in the 2.4
GHz and 5 GHz license-free ISM bands. However, WLANs do not offer the kind of
mobility, which mobile systems do.
1.2 Broadband Wireless Communication System
The explosive growth of the Internet, and the success of 2G and 3G systems
together with WLANs have had a profound impact on our perception of
communication. First of all, the vast majority of users now believe in the new notion
of “always on” communication. We are now living in the era of ubiquitous
connectivity or “communication anytime, anywhere, and with anything”.
Figure 1.2: Current and Future Wireless Communication Systems [2], [4], [5].
4
Secondly, the concept of broadband communication has caught on very well. As
fibre penetrates closer to the end-user environment (Fibre To The Home/Curb/X,
FTTH/C/X), wired transmission speeds will continue to rise. Transmission speeds
such at 100 Mbps (Fast Ethernet) are now beginning to reach homes. The demand to
have this broadband capacity also wirelessly has put pressure on wireless
communication systems to increase both their transmission capacity, as well as their
coverage.
In general there is a trade off between coverage and capacity. Figure 1.2
shows the relationship between some of the various standards (present and future), in
terms of mobility (coverage), and capacity. For instance, the cell size of Wireless
Personal Area Networks (WPANs) is typically a few metres (picocell), while their
transmission rates may reach several tens of Mbps. On the other hand 2G (e.g. GSM),
and 3G (e.g. Universal Mobile Telecommunication System (UMTS) and the
International Mobile Telecommunications (IMT2000)) systems have cells that extend
several kilometres, but have data rates limited to less than 2 Mbps. Therefore, as
mobile communication systems seek to increase capacity, and wireless data systems
seek to increase coverage, they will both move towards convergence. A case in point
is the IEEE 802.16, otherwise known as WiMAX, which appears to lend weight to the
notion of convergence, as shown in Figure 1.2. WiMAX seeks to provide high-bit rate
mobile services using frequencies between 2 – 11 GHz. In addition, WiMAX also
aims to provide Fixed Wireless Access (FWA) at bit-rates in the excess of 100 Mbps
and at higher frequencies between 10 – 66 GHz [6].
1.3 Challenges of Broadband Wireless Access Network
Figure 1.3 illustrates the configuration of narrowband wireless access systems
(e.g. GSM) as we know them today. The central office handles call processing and
switching, while the Base Stations (BS) act as the radio interfaces for the Mobile
Units (MU) or Wireless Terminal Units (WTU). The BSs may be linked to the central
office through either analogue microwave links or digital fibre optic links. Once the
baseband signals are received at the BS, they are processed and modulated onto the
appropriate carrier. The radius covered by the signal from the BS is the cell radius.
5
All the MU/WTU within the cell, share the radio frequency spectrum. WLANs are
configured in a similar fashion, with the radio interface called the Radio Access Point
(RAP).
Figure 1.3: Components of Narrowband Wireless Access Network.
In general, low carrier frequencies offer low bandwidth. Therefore, part of the
reason why narrowband wireless access systems (e.g. 2G) offer limited capacity is
because they operate at low frequencies. For instance GSM operates at frequencies
around 900 or 1800 MHz with 200 kHz allocated frequency spectrum. UMTS
operates at frequencies around 2 GHz and has 4 MHz allocated bandwidth [7].
However, there is also stiff competition for frequency spectrum among the many
wireless communication systems using carrier frequencies below 6 GHz. These
include radio and TV broadcasts, and systems for (vital) communication services such
as airports, police and fire, amateur radio users, wireless LANs, and many others.
Low frequencies allow for low cost radio front-ends (in the BS and the MU/WTU). In
addition, the efficiency of RF active devices (transistors) is higher at low frequencies,
than at high frequencies. For instance, at millimetre wave frequencies the efficiency
of active devices can be as low as 30 % [9]. Therefore, the low-power consumption
advantage of systems operating at low frequencies is quite significant. Furthermore,
low-frequency RF signals allow for larger cells, due to the longer reach of the radio
6
waves. The larger cells enable high mobility, but lead to poor spectrum efficiency,
since the spectrum is shared by all MUs/WTUs operating within the cell [7].
Therefore, one natural way to increase capacity of wireless communication
systems is to deploy smaller cells (micro- and pico-cells). This is generally difficult to
achieve at low-frequency microwave carriers, but by reducing the radiated power at
the antenna, the cell size may be reduced somewhat. Pico-cells are also easier to form
inside buildings, where the high losses induced by the building walls help to limit the
cell size. In contrast, the high propagation losses, which radio waves experience at
mmwave frequencies, together with the line-of-site requirements, help to form small
cells. Another way to increase the capacity of wireless communication systems is to
increase the carrier frequencies, to avoid the congested ISM band frequencies. Higher
carrier frequencies offer greater modulation bandwidth, but may lead to increased
costs of radio front-ends in the BSs and the MUs/WTUs.
Smaller cell sizes lead to improved spectral efficiency through increased
frequency reuse. But, at the same time, smaller cell sizes mean that large numbers of
BSs or RAPs are needed in order to achieve the wide coverage required of ubiquitous
communication systems. Furthermore, extensive feeder networks are needed to
service the large number of BSs/RAPs. Therefore, unless the cost of the BSs/RAPs,
and the feeder network are significantly low, the system-wide installation and
maintenance costs of such systems would be rendered prohibitively high. This is
where Radio-over-Fibre (RoF) technology comes in. It achieves the simplification of
the BSs/RAPs (referred to as Remote Antenna Units – RAUs) through consolidation
of radio system functionalities at a centralised headend, which are then shared by
multiple RAUs. In addition, a further reduction in system costs may be achieved if
low-cost multimode fibres are used in the extensive feeder network [10].
Therefore, for broadband wireless communication systems to offer the needed
high capacity, it appears inevitable to increase the carrier frequencies and to reduce
cell sizes. This is evident from the new standards in the offing, which are aiming to
use mm-waves. For example the recently formed IEEE 802.15 WPAN standard Task
Group 3c is aiming to use the unlicensed mm-wave bands between 57 - 64 GHz for
7
very-high-speed short-range communication offering up-to 2 Gbps [7]. The IEEE
802.16 (WiMAX) standard specifies frequencies between 10–66 GHz for the first/last
mile Fixed Wireless Access (FWA). A summary of the operating frequencies for
some of the current and future (broadband) wireless systems is given in Table 1.2.
Table 1.2: Frequencies for Wireless Broadband Communications Frequency Wireless Systems
2 GHz UMTS / 3G Systems
2.4 GHz IEEE 802.11 b/g WLAN
5 GHz IEEE 802.11 a WLAN
2 – 11 GHz IEEE 802.16 WiMAX
17/19 GHz Indoor Wireless (Radio) LANs
28 GHz Fixed wireless access – Local point to Multipoint (LMDS)
38 GHz Fixed wireless access, Picocellular
58 GHz Indoor wireless LANs
57 – 64 GHz IEEE 802.15 WPAN
10 – 66 GHz IEEE 802.16 - WiMAX
1.4 Radio over Fibre Technology
1.4.1. What is RoF
Radio-over-Fibre (RoF) technology entails the use of optical fibre links to
distribute RF signals from a central location (headend) to Remote Antenna Units
(RAUs). In narrowband communication systems and WLANs, RF signal processing
functions such as frequency up-conversion, carrier modulation, and multiplexing, are
performed at the BS or the RAP, and immediately fed into the antenna. RoF makes it
possible to centralise the RF signal processing functions in one shared location
(headend), and then to use optical fibre, which offers low signal loss (0.3 dB/km for
1550 nm, and 0.5 dB/km for 1310 nm wavelengths) to distribute the RF signals to the
RAUs, as shown in Figure 1.4. By so doing, RAUs are simplified significantly, as
they only need to perform optoelectronic conversion and amplification functions. The
centralisation of RF signal processing functions enables equipment sharing, dynamic
allocation of resources, and simplified system operation and maintenance. These
benefits can translate into major system installation and operational savings [8],
8
especially in wide-coverage broadband wireless communication systems, where a
high density of BS/RAPs is necessary as discussed above.
Figure 1.4: The RoF Concept.
One of the pioneer RoF system implementations is depicted in Figure 1.5.
Such a system may be used to distribute GSM signals, for example. The RF signal is
used to directly modulate the laser diode in the central site (headend). The resulting
intensity modulated optical signal is then transported over the length of the fibre to the
BS (RAU). At the RAU, the transmitted RF signal is recovered by direct detection in
the PIN photodetector. The signal is then amplified and radiated by the antenna. The
uplink signal from the MU is transported from the RAU to the headend in the same
way. This method of transporting RF signals over the fibre is called Intensity
Modulation with Direct Detection (IM-DD), and is the simplest form of the RoF link.
While Figure 1.5 shows the transmission of the RF signal at its frequency, it is
not always necessary to do that. For instance, a Local Oscillator (LO) signal, if
available, may be used to down-convert the uplink carrier to an IF in the RAU. Doing
so would allow for the use of low-frequency components for the up-link path in the
RAU – leading to system cost savings. Instead of placing a separate LO in the RAU, it
may be transported from the headend to the RAU by the RoF system. Once available
at the RAU, the LO may then be used to achieve down-conversion of the uplink
9
signals. This results in a much simpler RAU. In this configuration, the downlink
becomes the crucial part of the RoF since it has to transport high-frequency signals.
The transportation of high-frequency signals is more challenging because it require
high frequency components, and large link bandwidth. This means that high-
frequency signals are more susceptible to transmitter, receiver, and transmission link
signal impairments.
Figure 1.5: 900MHz Fibre-Radio System [9].
Apart from IM-DD, other methods, which involve signal frequency up-
conversion in addition to distribution, can also be used.
1.4.2. Advantages of RoF
Some of the advantages and benefits of the RoF technology compared with
electronic signal distribution are given below.
A. Low Attenuation Loss
Electrical distribution of high-frequency microwave signals either in free
space or through transmission lines is problematic and costly. In free space, losses due
to absorption and reflection increase with frequency [5]. In transmission lines,
impedance rises with frequency as well, leading to very high losses [11]. Therefore,
distributing high frequency radio signals electrically over long distances requires
expensive regenerating equipment. As for mm-waves, their distribution via the use of
transmission lines is not feasible even for short distances. The alternative solution to
this problem is to distribute baseband signals or signals at low intermediate
frequencies (IF) from the switching centre (headend) to the BS [1]. The baseband or
10
IF signals are up-converted to the required microwave or mm-wave frequency at each
base station, amplified and then radiated. This system configuration is the same as the
one used in the distribution of narrowband mobile communication systems shown in
Figure 1.3. Since, high performance LOs would be required for up-conversion at each
base station, this arrangement leads to complex base stations with tight performance
requirements. However, since optical fibre offers very low loss, RoF technology can
be used to achieve both low-loss distribution of mm-waves, and simplification of
RAUs at the same time. Commercially available standard Single Mode Fibres (SMFs)
made from glass (silica) have attenuation losses below 0.2 dB/km and 0.5 dB/km in
the 1550 nm and the 1300 nm windows, respectively. Polymer Optical Fibres (POFs),
a more recent kind of optical fibre exhibits higher attenuation ranging from 10 – 40
dB/km in the 500 – 1300 nm regions [12], [13]. These losses are much lower than
those encountered in, say coaxial cable, whose losses are higher by three orders of
magnitude at higher frequencies. For instance, the attenuation of a 1/2 inch coaxial
cable (RG-214) is >500 dB/km for frequencies above 5 GHz [14]. Therefore, by
transmitting microwaves in the optical form, transmission distances are increased
several folds and the required transmission powers reduced greatly.
B. Large Bandwidth
Optical fibres offer enormous bandwidth. There are three main transmission
windows, which offer low attenuation, namely the 850 nm, 1310 nm, and 1550 nm
wavelengths. For a single SMF optical fibre, the combined bandwidth of the three
windows is in the excess of 50 THz [15]. However, today’s state-of-the-art
commercial systems utilize only a fraction of this capacity (1.6 THz). But
developments to exploit more optical capacity per single fibre are still continuing. The
main driving factors towards unlocking more and more bandwidth out of the optical
fibre include the availability of low dispersion (or dispersion shifted) fibre, the
Erbium Doped Fibre Amplifier (EDFA) for the 1550 nm window, and the use of
advanced multiplex techniques namely Optical Time Division Multiplexing (OTDM)
in combination with Dense Wavelength Division Multiplex (DWDM) techniques. The
enormous bandwidth offered by optical fibres has other benefits apart from the high
capacity for transmitting microwave signals. The high optical bandwidth enables high
speed signal processing that may be more difficult or impossible to do in electronic
11
systems. In other words, some of the demanding microwave functions such as
filtering, mixing, up- and down-conversion, can be implemented in the optical domain
[16]. For instance, mm-wave filtering can be achieved by first converting the
electrical signal to be filtered into an optical signal, then performing the filtering by
using optical components such as the Mach Zehnder Interferometer (MZI) or Fibre
Bragg Gratings (FBG), and then converting the filtered signal back into electrical
form [17]. Furthermore, processing in the optical domain makes it possible to use
cheaper low bandwidth optical components such as laser diodes and modulators, and
still be able to handle high bandwidth signals [17]. The utilization of the enormous
bandwidth offered by optical fibres is severely hampered by the limitation in
bandwidth of electronic systems, which are the primary sources and receivers of
transmission data. This problem is referred to as the “electronic bottleneck”. The
solution around the electronic bottleneck lies in effective multiplexing. OTDM and
DWDM techniques mentioned above are used in digital optical systems. In analogue
optical systems including RoF technology, Sub-Carrier Multiplexing (SCM) is used to
increase optical fibre bandwidth utilization. In SCM, several microwave subcarriers,
which are modulated with digital or analogue data, are combined and used to
modulate the optical signal, which is then carried on a single fibre [19], [20]. This
makes RoF systems cost-effective.
C. Immunity to Radio Frequency Interference
Immunity to ElectroMagnetic Interference (EMI) is a very attractive property
of optical fibre communications, especially for microwave transmission. This is so
because signals are transmitted in the form of light through the fibre. Because of this
immunity, fibre cables are preferred even for short connections at mm-waves. Related
to EMI immunity is the immunity to eavesdropping, which is an important
characteristic of optical fibre communications, as it provides privacy and security.
D. Easy Installation and Maintenance
In RoF systems, complex and expensive equipment is kept at the headend,
thereby making the RAUs simpler. For instance, most RoF techniques eliminate the
need for a LO and related equipment at the RAU. In such cases a photodetector, an
RF amplifier, and an antenna make up the RAU. Modulation and switching equipment
is kept in the headend and is shared by several RAUs. This arrangement leads to
12
smaller and lighter RAUs, effectively reducing system installation and maintenance
costs. Easy installation and low maintenance costs of RAUs are very important
requirements for mm-wave systems, because of the large numbers of the required
RAUs. In applications where RAUs are not easily accessible, the reduction in
maintenance requirements leads to major operational cost savings [8], [11]. Smaller
RAUs also lead to reduce environmental impact.
E. Reduced Power Consumption
Reduced power consumption is a consequence of having simple RAUs with
reduced equipment. Most of the complex equipment is kept at the centralised headend.
In some applications, the RAUs are operated in passive mode. For instance, some 5
GHz Fibre-Radio systems employing pico-cells can have the RAUs operate in passive
mode [21]. Reduced power consumption at the RAU is significant considering that
RAUs are sometimes placed in remote locations not fed by the power grid.
F. Multi-Operator and Multi-Service Operation
RoF offers system operational flexibility. Depending on the microwave
generation technique, the RoF distribution system can be made signal-format
transparent. For instance the Intensity Modulation and Direct Detection (IM-DD)
technique can be made to operate as a linear system and therefore as a transparent
system. This can be achieved by using low dispersion fibre (SMF) in combination
with pre-modulated RF subcarriers (SCM). In that case, the same RoF network can be
used to distribute multi-operator and multi-service traffic, resulting in huge economic
savings [8].
G. Dynamic Resource Allocation
Since the switching, modulation, and other RF functions are performed at a
centralized headend, it is possible to allocate capacity dynamically. For instance in a
RoF distribution system for GSM traffic, more capacity can be allocated to an area
(e.g. shopping mall) during peak times and then re-allocated to other areas when
offpeak (e.g. to populated residential areas in the evenings). This can be achieved by
allocating an optical wavelength through Wavelength Division Multiplexing (WDM)
as need arises. Allocating capacity dynamically as need for it arises obviates the
requirement for allocating permanent capacity, which would be a waste of resources
13
in cases where traffic loads vary frequently and by large margins [8]. Furthermore,
having the centralised headend facilitates the consolidation of other signal processing
functions such as mobility functions, and macro diversity transmission.
1.4.3. Limitations of RoF
Since RoF involves analogue modulation, and detection of light, it is
fundamentally an analogue transmission system. Therefore, signal impairments such
as noise and distortion, which are important in analogue communication systems, are
important in RoF systems as well. These impairments tend to limit the Noise Figure
(NF) and Dynamic Range (DR) of the RoF links. DR is a very important parameter
for mobile (cellular) communication systems such as GSM because the power
received at the BS from the MUs varies widely (e.g. 80 dB [8]). That is, the RF power
received from a MU, which is close to the BS, can be much higher than the RF power
received from a MU, which is several kilometres away, but within the same cell.
The noise sources in analogue optical fibre links include the laser’s Relative
Intensity Noise (RIN), the laser’s phase noise, the photodiode’s shot noise, the
amplifier’s thermal noise, and the fibre’s dispersion. In Single Mode Fibre (SMF)
based RoF, systems, chromatic dispersion may limit the fibre link lengths and may
also cause phase de-correlation leading to increased RF carrier phase noise [5]. In
Multi-Mode Fibre based RoF systems, modal dispersion severely limits the available
link bandwidth and distance. It must be stated that although the RoF transmission
system itself is analogue, the radio system being distributed need not be analogue as
well, but it may be digital (e.g. WLAN, UMTS), using comprehensive multi-level
signal modulation formats such as xQAM, or Orthogonal Frequency Division
Multiplexing (OFDM).
1.5 Integrating Wireless and Fibre Optics
For the future provision of broadband, interactive and multimedia services
over wireless media, current trends in cellular networks - both mobile and fixed are 1)
to reduce cell size to accommodate more users and 2) to operate in the
microwave/millimetre wave (mm-wave) frequency bands to avoid spectral congestion
14
in lower frequency bands. It demands a large number of base stations (BSs) to cover a
service area, and cost-effective BS is a key to success in the market. This requirement
has led to the development of system architecture where functions such as signal
routing/processing, handover and frequency allocation are carried out at a central
control station (CS), rather than at the BS. Furthermore, such a centralized
configuration allows sensitive equipment to be located in safer environment and
enables the cost of expensive components to be shared among several BSs. An
attractive alternative for linking a CS with BSs in such a radio network is via an
optical fibre network, since fibre has low loss, is immune to EMI and has broad
bandwidth. The transmission of radio signals over fibre, with simple optical to
electrical conversion, followed by radiation at remote antennas, which are connected
to a central CS, has been proposed as a method of minimizing costs. The reduction in
cost can be brought about in two ways. Firstly, the remote antenna BS or radio
distribution point needs to perform only simple functions, and it is small in size and
low in cost. Secondly, the resources provided by the CS can be shared among many
antenna BSs.
To be specific, the RoF network typically comprises a central CS, where all
switching, routing, medium access control (MAC) and frequency management
functions are performed, and an optical fibre network, which interconnects a large
number of functionally simple and compact antenna BSs for wireless signal
distribution. The BS has no processing function and its main function is to convert
optical signal to wireless one and vice versa. Since RoF technology was first
demonstrated for cordless or mobile telephone service in 1990 [22], a lot of research
efforts have been made to investigate its limitation and develop new, high
performance RoF technologies. Their target applications range from mobile cellular
networks [23]-[25], wireless local area network (WLAN) at mm-wave bands [26],
broadband wireless access networks [27]-[30] to road vehicle communication (RVC)
networks for intelligent transportation system (ITS) [31]-[33]. Due to the simple BS
structure, system cost for deploying infrastructure can be dramatically reduced
compared to other wireline alternatives. In addition to the advantage of potential low
cost, RoF technology has the further a benefit of transferring the RF signal to and
from a CS that can allow flexible network resource management and rapid response to
15
variations in traffic demand due to its centralized network architecture. In summary,
some of its important characteristics are described below [8]:
1. The system control functions, such as frequency allocation, modulation and
demodulation scheme, are located within the CS, simplifying the design of the BS.
The primary functions of the BSs are optical/RF conversion, RF amplification,
and RF/optical conversion.
2. This centralized network architecture allows a dynamic radio resource
configuration and capacity allocation. Moreover, centralized upgrading is also
possible.
3. Due to simple BS structure, its reliability is higher and system maintenance
becomes simple.
4. In principle, optical fibre in RoF is transparent to radio interface format
(modulation, radio frequency, bit rate and so on) and protocol. Thus, multiple
services on a single fibre can be supported at the same time.
5. Large distances between the CS and the BS are possible.
On the other hand, to meet the explosive demands of high-capacity and
broadband wireless access, millimetre-wave (mm-wave) radio links (26ñ100 GHz) are
being considered to overcome bandwidth congestion in microwave bands such as 2.4
or 5 GHz for application in broadband micro/picocellular systems, fixed wireless
access, WLANs, and ITSs [34]-[37]. The larger RF propagation losses at these bands
reduce the cell size covered by a single BS and allow an increased frequency reuse
factor to improve the spectrum utilization efficiency. Recently, considerable attention
has been paid in order to merge RoF technologies with mm-wave band signal
distribution [38]-[44]. The system has a great potential to support cost-effective and
high capacity wireless access. The distribution of radio signals to and from BSs can
be either mm-wave modulated optical signals (RF-over-fibre), or lower frequency
subcarriers (IF-over-fibre). Signal distribution as RF-over-fibre has the advantage of a
simplified BS design but is susceptible to fibre chromatic dispersion that severely
16
limits the transmission distance [45]. In contrast, the effect of fibre chromatic
dispersion on the distribution of intermediate-frequency (IF) signals is much less
pronounced, although antenna BSs implemented for RoF system incorporating IF-
over-fibre transport require additional electronic hardware such as a mm-wave
frequency local oscillator (LO) for frequency up and down conversion. These research
activities fuelled by rapid developments in both photonic and mm-wave technologies
suggest simple BSs based on RoF technologies will be available in the near future.
However, while great efforts have been made in the physical layer, little attention has
been paid to upper layer architecture. Specifically, centralized architecture of RoF
networks implies the possibility that resource management issues in conventional
wireless networks could be efficiently addressed. As a result, it is required to
reconsider conventional resource management schemes in the context of RoF
networks.
1.6 The Research Aim and Contributions
The aim of this work is to enhance the multi radios over fibre architecture in
the following aspects:
A. Application Controlled Handover This work provides a solution to enhance the radio interface selection for the
mobile user. The radio interface selection criterion is based on the application type
such as HTTP, FTP and video conferencing and communications. The proposed
enhanced handover mechanism is cross layer design according to the current request
and session of the device application. Applying the proposed approach leads to better
radio interface selection and more network efficiency as shown in the achieved results
in Chapter three.
B. Cross Layer Information Exchange The second major goal of this research is to introduce the cross layer concept
to the wired wireless converged standard. According to this concept, the Application,
Transport, MAC and PHY layers exchange application type information to initiate
radio interface suitable for application requested by user. This concept enhances the
17
standard network convergence and integration. If the user has lower battery, the
handover mechanism will initiate for suitable interface for lower power consumption.
C. Mesh Formation and Scalability
Mesh formation for multi radio over optical network integration has been
introduced in Chapter four. Wireless gives you mobility, optical fibre is more secure,
and have terahertz of bandwidth. WiFi and WiMAX have been used for mesh
connectivity to router traffic through mesh routers with a functionality of radio access
units.
D. Power Management
Power management in wireless radios has been studied in this work. A simple
model is based on controlling the transmitter power to reach the minimum required
power at the receiver. The data sender and the data receiver devices exchange control
messages before and during data transmission. These control messages are used to
maintain the transmission power in the sender device to the minimum. This approach
is simple and efficient comparing to other power management approaches. This
approach minimises the interference and reduces the general power consumption in
the network.
E. Application Priority Assignment for BWSN Furthermore, application priority based on the important of medical data has been
proposed in order to enhance the wireless sensor network model. This simulation
model is used to emulate the real-life wireless sensors, where the medical data is
related upon the application type. In the proposed model, the device applications are
switched between different applications and virtual data queues are prioritised. An
intelligent application priority assignment mechanism has been described, developed
and proved by simulation that end-to-end delay is reduced and the network throughput
is increased comparing to the normal standard of BWSN as presented in Chapter five.
F. Mobility Strategies for IAPAM enabled BWSN
An IEEE 802.15.4 based IAPAM enabled BWSN is composed at variable speeds
(5m/s, 10 m/s and 20 m/s) which depicts a mobile patient at hospital or home. In
18
proposed scenarios, the mobile node uses different mobility patterns and their impact
on the performance over the mobile BWSN.
1.7 Applications and the Future Networks
RoF technology is generally unsuitable for system applications, where high
Spurious Free Dynamic Range (SFDR = maximum output signal power for which the
power of the third-order inter-modulation product is equal to the noise floor) is
required, because of the limited DR. This is especially true of wide coverage mobile
systems such as GSM, where SFDR of > 70 dB (outdoor) are required. However,
most indoor applications do not require high SFDR. For instance, the required
(uplink) SFDR for GSM reduces from >70 dB to about 50 dB for indoor applications.
Therefore, RoF distribution systems can readily be used for in-building (indoor)
distribution of wireless signals of both mobile and data communication (e.g. WLAN)
systems. In this case the RoF system becomes a Distributed Antenna System (DAS).
For high frequency applications such as WPAN, the cell size will be small due to high
losses through the walls, bringing the advantages of RoF discussed above. The in-
building fibre infrastructure may then be used for both wired and wireless applications.
Using MMF or indeed POF instead of SMF fibres to feed the RAUs may further
reduce system installation and maintenance costs, especially for in-door applications.
In-building data communications LANs are often based on MMF.
RoF systems are also attractive for other present and future applications where
high SFDR is not required. For instance, UMTS MUs are required to control their
transmitter power so that equal power levels are received at the BS. Thus, UMTS does
not need the high SFDR required in GSM, so that RoF distribution systems may be
used for both indoor and outdoor UMTS signal distribution [8]. Another application
area is in Fixed Wireless Access (FWA) systems, such as WiMAX, where RoF
technology may be used to optically transport signals over long distances bringing the
significantly simplified RAUs closer to the end user, from where wireless links help
to achieve broadband first/last mile access, in a cost effective way.
High data rate and quality of service QoS are crucial for video, audio and
Internet interaction. Furthermore, future applications such as health care and personal
19
security applications should be considered in the WPAN HR protocol with special
requirements for power consumption and data security. New approach for more
efficient Service Discovery (SD) algorithm is required in order to enhance the
expanding WPAN networks. Several approaches of service discovery could be
applied to the WPAN and cross-layered with the MAC HR protocol.
Internet dynamic access and smart house applications could be adapted by the
WPAN protocol. Internet gateway interaction with WPAN could be developed in the
aspect of access point to access point handover, where the personal device could
maintain external connectivity when it is relocated.
Cross Layer interaction could be applied in order to exchange parameters,
which belong to various layers in the wireless device (from PHY to APP layers).
Lower layers such as PHY/MAC and D-LINK could exchange some parameters with
higher layers such as the TRANS and APP layers. These parameters provide more
information about the user and the device/network status.
1.8 Objectives and Research Methodology
This research is based on the following steps:
1. Review of the wireless and fibre optics standard and the related publications about
their convergence and integration protocols.
2. Intensive study and assessment of the potential weakness in the optical wireless
hybrid standard and suggest a network architecture solution to enhance FMC.
3. Modelling and analysis of the application controlled protocol performance in
order to prove the improved network design.
4. Introducing power management scheme and mesh formation in wireless domain to
enhance application controlled proposed converged next generation network.
20
5. Simulate the proposed approach and compare it with the standard optical wireless
integrated network.
6. Application priority assignment for BWSN to enhance the transmission
mechanism of patient’s critical data.
7. Verify the models and analyze the mobility patterns for biomedical wireless
sensor networks.
1.8.1. Simulation Environment
In this thesis, OPNET [46] and JSIM [47] are used to implement and study
optical wireless integrated environment and wireless sensor network design. Prior to
start implementing the model, a training period is required in order to build some
skills in simulation tools. Some simple models are implemented in OPNET and
improved to accomplish the final design. Furthermore, simulators documentations
help files and the supplied tutorial have been applied to enhance the simulation
experience.
1.8.2. Model Design
Building a custom model using OPNET node simulates the application
controlled handover model. Behaviour of all devices in the network is implemented in
this design. The node model is mapped to a process model, which is designed using
C/C++ languages and some built-in OPNET library functions. JSIM is used for
simulations for BWSN. Based on same theory of application controlled handover,
application priority assignment is developed in JSIM to enhance the proficiency of
standard BWSN.
1.8.3. Comparison of Simulation Results
The proposed designs are compared with the original standards in order to
show the enhancement achieved in the proposed approaches. The new designs are
21
based on introducing the user applications and improvement of BWSN. The
simulation results are presented in Chapter Three, Chapter Four, Chapter Five and
Chapter Six respectively.
1.9 The Thesis Structure
This section presents the organization of this thesis, where Chapter One
presents the introduction of the work with some background about the wireless
communication systems and radio over fibre technologies. For none-expert readers,
this chapter is important to understand the rest of this thesis.
Chapter Two presents the relevant literature review, where previous work and
several papers in the related field have been presented.
Chapter Three is dealing with the proposal of application controlled handover
scheme, enhancement of the integrated network protocol by proposing a new
capability of switching criteria of radio interfaces. In this chapter, the integrated
hybrid network based on multiple radio interfaces and RoF technology are described.
Then, the smart handover scheme presented, developed and simulated. Finally, the
comparative results are discussed in details.
In Chapter Four, the power management scheme is presented. The power
controlled scheme is implemented on top of application controlled handover to
enhance the network performance in terms of power and data transmission. A WiFi
and WiMAX mesh access units (Mesh Routers) are also deployed. The proposed
approach uses much lower power and higher throughput. The mesh based power and
application controlled handover concept in the optical wireless integrated network is
described in detail in this chapter and simulation scenario using OPNETTM modelor is
implemented and results have shown the network improvements.
In Chapter Five, description of the BWSN simulation model and JSIM
simulation environment are presented. In this chapter, intelligent application priority
assignment mechanism is described with the control and switching module. Virtual
queues and new set of parameters proposed in superframe of standard BWSN. Priority
queue collects the critical data and triggers the urgent transmission for patient’s timely
22
monitoring. The scheme is proposed, discussed, developed and simulated and
compared in this chapter.
In Chapter six, description of the mobility models for IAPAM enabled BWSN
has been presented. In this chapter, a successful study of variable mobility has been
introduced over IAPAM enabled BWSN. The random, controlled and directed
mobility schemes are discussed, deployed and simulated and discussed in this chapter.
Chapter Seven presents the research conclusions and the future work.
Furthermore, summary of the challenges and solutions, which are presented in this
thesis are included in Chapter Seven.
At the end of the thesis, a list of the published and submitted papers is
presented.
1.10 References
[1] ITU, “World Telecommunication Development Report 2002: Reinventing
Telecoms”, March, 2002.
[2] Internal Ericsson.
http://www.ericsson.com/thinkingahead/themes?themes=#e_new_generation_ip_n
etworking_1523462065_c.
[3] The Industrial Wireless Book, “De-mystifying IEEE 802.11 for Industrial
Wireless LANs”, available online at http://wireless.industrial-networking.com,
May 2005.
[4] S. Ohmori, “The Future Generations of Mobile Communications Based on
Broadband Access Technologies”, IEEE Communications Magazine, pp. 134-142,
December 2000.
[5] D. Novak, “Fiber Optics in Wireless Applications”, OFC 2004 Short Course
Notes, 2004.
[6] IEEE 802.16 Working Group on Broadband Wireless Access Standards,
Developing the IEEE 802.16 WirelessMAN® Standard for Wireless Metropolitan
Area Networks”, available online at http://ieee802.org/16/, May 2005.
23
[7] IEEE 802.15 Working Group for Wireless Personal Area Networks, “IEEE 802.15
WPAN Millimeter Wave Alternative PHY Task Group 3c (TG3c)”, available
online at http://www.ieee802.org/15/pub/TG3c.html, May 2005.
[8] D. Wake, “Radio over Fiber Systems for Mobile Applications in Radio over Fiber
Technologies for Mobile Communications Networks”, H. Al-Raweshidy and S.
Komaki, edited by Artech House Inc, USA, 2002.
[9] D. M. Pozar, “Microwave and RF Design of Wireless Systems”, John Wiley
Sons, Inc., 2001.
[10] T. Koonen, A. Ng’oma, P. F. M. Smulders, H. P. A. vd. Boom, I. Tafur Monroy,
and G. D. Khoe, “In-House networks using Multimode Polymer Optical Fiber for
broadband wireless services”, Photonic Network Communications, Vol. 5, No. 2,
177-187, Kluwer, 2003.
[11] J. J. O’Reilly, P. M. Lane, and M. H. Capstick, “Optical Generation and
Delivery of Modulated mm-waves for Mobile Communications”, in Analogue
Optical Fibre Communications, B. Wilson, Z. Ghassemlooy, and I. Darwazeh, ed.
The Institute of Electrical Engineers, London, 1995.
[12] Y. Koike, “POF Technology for the 21st Century”, in Proceedings of the Plastic
Optical Fibres (POF) Conference, pp. 5-8, 2001.
[13] Y. Watanabe, “Current Status of Perfluorinated GI-POF and 2.5 Gbps Data
Transmission over it”, in Proceedings of OFC ‘03, USA, pp. 12-13, 2003.
[14] Wake, and K. Beachman, “A Novel Switched Fibre Distributed Antenna
System”, in Proceedings of European Conference on Optical Communications
(ECOC’04), Vol. 5, pp. 132–135, 2004.
[15] D. K. Mynbaev, L. L. Scheiner, “Fiber Optic Communications Technology”,
Prentice Hall, New Jersey, 2001.
[16] J. Capmany, B. Ortega, D. Pastor, and S. Sales, “Discrete-Time Optical
Processing of Microwave Signals”, JLT, Vol. 23, No. 2, 703 - 723, 2005.
[17] B. Carbon, V. Girod, and G. Maury, “Optical Generation of Microwave
Functions”, in Proceedings of Workshop on Microwave Photonics for Emission
and Detection of Broadband Communication Signals, Louvain-la-Nueve, Belgium,
2001.
24
[18] G. Maury, A. Hilt, T. Berceli, B. Cabon, and A. Vilcot, “Microwave Frequency
Conversion Methods by Optical Interferometer and Photodiode”, IEEE Trans. On
Microwave Theory and Techniques, Vol. 45, No. 8, 1481 - 1485, 1997.
[19] D. Wake, S. Dupont, J-P. Vilcot, and A. J. Seeds, “32-QAM Radio
Transmission Over Multimode Fibre Beyond the Fibre Bandwidth”, in
Proceedings of the IEEE International Topical Meeting on Microwave
Photonics MWP, 2001.
[20] D. Wake, S. Dupont, C. Lethien, J-P. Vilcot, and D. Decoster, “Radiofrequency
Transmission over Multimode Fibre for Distributed Antenna System
Applications”, Electronic Letters, Vol. 37, No. 17, pp. 1087–1089, 2001.
[21] C. Liu, A. Seeds, J. Chadha, P. Stavrinou, G. Parry, M. Whitehead, A. Krysa,
and J. Roberts, “Bi-Directional Transmission of Broadband 5.2 GHz Wireless
Signals Over Fibre Using a Multiple-Quantum-Well Asymetric Fabry-Perot
Modulator/Photodetector”, in Proceedings of the Optical Fiber Communications
(OFC) Conference, Vol. 2, pp. 738–740, 2003.
[22] A. J. Cooper, Fiber/Radio for the provision of cordless/mobile telephony
services in the access network, Electron. Letter, vol. 26, no. 24, pp. 2054-2056,
Nov. 1990.
[23] T. S. Chu and M. J. Gans, Fiber Optic Microcellular Radio, IEEE Trans. Veh.
Technology, vol. 40, no. 3, pp. 599-606, August 1991.
[24] H. Kim, J. M. Cheong, C. Lee, and Y. C. Chung, ìPassive Optical Network for
Microcellular CDMA Personal Communication Service, IEEE Photonics Tech.
Letter, vol. 10, no. 11, pp. 1641-1643, Nov 1998.
[25] J. S. Yoon, M. H. Song, S. H. Seo, Y. S. Son, J. M. Cheong and Y. K. Jhee, A
High-Performance Fiber-Optic Link for cdma2000 Cellular Network, IEEE
Photon. Technol. Letter, vol. 14, no. 10, pp. 1475-1477, October 2002.
[26] K. Kojucharow, M. Sauer, H. Kaluzni, D. Sommer, F. Poegel, W. Nowak, and
D. Ferling, Simultaneous Electro-optical Upconversion, Remote Oscillator
Generation, and Air Transmission of Multiple Optical WDM Channels for a 60-
GHz High-Capacity Indoor System, IEEE Trans. Microwave Theory Technology,
vol. 47, no. 12, pp. 2249-2255, December 1999.
[27] K. Kitayama and R. A. Grifin, Optical Downconversion from Millimeter-Wave
to IF-Band Over 50-km-Long Optical Fiber Link Using an Electroabsorption
25
Modulator, IEEE Photon. Technology. Letter, vol. 11, no. 2, pp. 287-289, Feb.
1999.
[28] K. Kitayama, A. St¨ohr, T. Kuri, R. Heinzelmann, D. Jager and Y. Takahashi,
An Approach to Single Optical Component Antenna Base Stations for Broad-
Band Millimeter-Wave Fiber-Radio Access Systems, IEEE Trans. Microwave
Theory Technology, vol. 48, pp. 2588-2594, December 2000.
[29] T. Kuri, K. Kitayama, and Y. Takahashi, ì60-GHz-Band Full-Duplex Radio-On-
Fiber System Using Two-RF-Port Electro-absorption Transceiver, IEEE Photon.
Technol. Letter, vol. 12, no. 4, pp. 419-421, April 2000.
[30] G. H. Smith, D. Novak, and C. Lim, A millimeter-wave full-duplex _ber-radio
star-tree architecture incorporating WDM and SCM, IEEE Photon. Technol.
Letter, vol. 10, pp. 1650-1652, Nov. 1998.
[31] H. Harada, K. Sato and M. Fujise, A Radio-on-Fiber Based Millimeter-Wave
Road-Vehicle Communication System by a Code Division Multiplexing Radio
Transmission Scheme, IEEE Transaction Intelligent Transport. Systems, vol. 2, no.
4, pp. 165-179, Dec. 2001.
[32] H. Harada, K. Sato and M. Fujise, A Radio-on-_ber Based Millimeter-wave
Road-vehicle Communication System For Future Intelligent Transport System, in
Proc. IEEE VTC 2001 Fall, vol. 4, pp. 2630-2634, Oct. 2001.
[33] Y. Okamoto, R. Miyamoto and M. Yasunaga, Radio-On-Fiber Access Network
Systems for Road- Vehicle Communication, in Proc. 2001 IEEE Intelligent
Transportation Sys. Conf., pp. 1050-1055, Aug. 2001.
[34] O. Andrisano, V. Tralli, and R. Verdone, ìMillimeter Waves for Short-Range
Multimedia Communication Systems, Proc. IEEE, vol. 86, no. 7, pp. 1383-1401,
July 1998.
[35] P. Smulders, ìExploiting the 60 GHz Band for Local Wireless Multimedia
Access: Prospects and Future Directions, IEEE Communications Magazine, pp.
140-147, Jan. 2002.
[36] J. R. Forrest, Communication Networks for the New Millennium, IEEE Trans.
Microwave Theory Technology, vol. 47, no. 12, pp. 2195-2201, Dec. 1999.
[37] F. Giannetti, M. Luise and R. Reggiannini, ìMobile and Personal
Communications in the 60 GHz Band: A Survey, Wireless Personal
Communication, pp. 207-243, Oct. 1999.
26
[38] H. Ogawa, D. Polifko and S. Banba, Millimeter-Wave Fiber Optics Systems for
Personal Radio Communication, IEEE Trans. Microwave Theory Tech., vol. 40,
no. 12, pp. 2285-2293, Dec 1992.
[39] R. Braun, G. Grosskopf, D. Rohde, and F. Schmidt, ìLow-Phase-Noise
Millimeter-Wave Generation at 64 GHz and Data Transmission Using Optical
Sideband Injection Locking, IEEE Photonics Technology Letter, vol. 10, no. 5, pp.
728-730, May. 1998.
[40] G. H. Smith and D. Novak, Broadband millimeter-wave fiber-radio network
incorporating remote up/down conversion, Microwave Symposium Digest, IEEE
MTT-S, vol. 3, pp. 1509-1512, Jun 1998.
[41] K. Kitayama and R. A. Grif_n, ìOptical Downconversion from Millimeter-
Wave to IF-Band Over 50-km-Long Optical Fiber Link Using an
Electroabsorption Modulator, IEEE Photon. Technology. Letter, vol. 11, no. 2, pp.
287-289, Feb. 1999.
[42] K. Kitayama, A. St¨ohr, T. Kuri, R. Heinzelmann, D. J¨ager and Y. Takahashi,
An Approach to Single Optical Component Antenna Base Stations for Broad-
Band Millimeter-Wave Fiber-Radio Access Systems, IEEE Trans. Microwave
Theory Technology, vol. 48, pp. 2588-2594, Dec. 2000.
[43] T. Kuri, K. Kitayama, and Y. Takahashi, ì60-GHz-Band Full-Duplex Radio-On-
Fiber System Using Two-RF-Port Electroabsorption Transceiver, IEEE Photonics.
Technol. Letter, vol. 12, no. 4, pp. 419ñ421, Apr. 2000.
[44] H. Harada, K. Sato and M. Fujise, A Radio-on-Fiber Based Millimeter-Wave
Road-Vehicle Communication System by a Code Division Multiplexing Radio
Transmission Scheme, IEEE Transaction of Intelligent Transportation System, vol.
2, no. 4, pp. 165-179, Dec. 2001.
[45] U. Gliese, S. Norskow, and T. N. Nielsen, Chromatic Dispersion in Fiber-Optic
Microwave and Millimeter-Wave Links, IEEE Trans. Microwave Theory
Technology, vol. 44, no. 10, pp. 1716-1724, Oct 1996.
[46] www.opnet.com (December 2010).
[47] www.j-sim.org (December 2010).
27
Chapter 2
Background and Literature Review ________________________________________________________
This chapter presents some basic background of Radio over Fibre technology,
a brief description of wireless technologies, handover issues and in depth literature
review for proposed handover technologies addressed in Chapters 3, 4 and 5. This
chapter constitutes of three major parts. The first part briefly covers basic optical fibre
transmission link and RoF technologies. The second part presents all related wireless
and mobile technologies; WLAN, WiMAX, WPAN (UWB, ZigBee) and UMTS,
while the third part unfolds the related work done in the relevant research area of
handover management schemes.
Therefore, this chapter will present the background and literature review in the
following fields:
• Radio over Fibre Technology
• Wireless Mobile Technologies
• Handover in Wireless Mobile Networks
2.1 Radio over Fibre Technologies
Wireless networks based on RoF technologies have been proposed as a
promising cost-effective solution to meet ever-increasing user bandwidth and wireless
demands. Since it was first demonstrated for cordless or mobile telephone service in
1990 [1], a lot of research has been carried out to investigate its limitation and
develop new and high performance RoF technologies. In this network a central station
(CS) is connected to numerous functionally simple BSs via an optic fibre. The main
function of BS is to convert optical signal to wireless one and vice versa. Almost all
processing including modulation, demodulation, coding, routing is performed at the
28
CS. That means, RoF networks use highly linear optic fibre links to distribute RF
signals between the CS and BSs.
Figure 2.1 presents a general RoF architecture. At a minimum, RoF link
consists of all the hardware required to impose an RF signal on an optical carrier, the
fibre optic link, and the hardware required to recover the RF signal from the carrier.
The optical carrier's wavelength is usually selected to coincide with either the 1.3 µm
at which standard single-mode fibre has minimum dispersion, or the 1.55 µm at which
its attenuation is minimum.
Figure 2.1: Radio over Fibre System [1].
2.1.1. Optical Transmission Link
In the first part of this section, a general optical transmission link, shown in
Figure 2.2, is briefly described for which it is assume that a digital pulse signal is
transmitted over optical fibre unless otherwise specified. The optical link consists of
an optical fibre, transmitter, receiver and two amplifiers.
29
Figure 2.2: Optical Transmission Link [1].
2.2 Wireless Mobile Technologies
Ever since the emergence of wireless networks as a ubiquitous solution that
increases connectivity in inaccessible areas, security has been a major concern. This
section presents different wireless networks by providing a better insight of their
functionalities and limitations. Furthermore a detailed analysis on the IEEE 802.x
wireless networks, with special focus on the WiFi and WiMAX is provided. As much
as the convergence of these two wireless networks is the main focus of this research,
other wireless networks such as UWB, Blutooth and ZigBee and private owned
mobile cellular networks are discussed as alternative technologies. The mobile
cellular technology is analyzed in terms of their suitability when mixed with either
WiFi or WiMAX in providing data, voice and video services to wireless networking
clients. The analysis of the converged WiFi and WiMAX highlights the protocols and
mechanism functionalities and limitations.
2.2.1. The IEEE 802.x wireless networks
The IEEE has established a variety of wireless standards and protocols, which
are defined within the 802 families. The 802.x family has a series of specifications for
different local area network (LAN) technologies [2]. In this section below brief
overview of IEEE 802.x wireless standards are presented.
A. IEEE 802.11 standards (WiFi)
The IEEE 802.11wireless standard, also known as Wireless Fidelity was
designed to provide flexibility and portability compared to the traditional wired
networks. These wireless standards specify the wireless interface between wireless
30
clients and the access point. The 802.11 standard is part of the IEEE 802 Networks
family. This standard defines how the wireless network handles the lower layers of
the OSI (Physical and Medium Access Control layers) [2] as they are different from
their wired networks counterpart. The PHY layer specifies all the details for
transmission and reception will the MAC layer set rules to determine how to access
the medium and transmission. The major aim for defining the 802.11 standard was to
resolve compatibility issues from wireless equipment manufacturers. The diagram
shows the detailed overview of the IEEE 802 network family
Figure 2.3: IEEE 802 in the OSI Layers [2].
Wireless devices are just like wired LAN devices in applications but differ in
their ability to operate and utilize wireless medium. Though the 802.11 based wireless
networks emerged from Ethernet networks their access methods are different. The
wired networks use Carrier Sense Multiple Access with Collision Detection
(CSMA/CD) that proved not to work well in wireless networks since they send and
receive data at the same time.
Therefore 802.11 networks use Carrier Sense Multiple Access with Collision
Avoidance (CSMA/CA), which improves the performance as it prevents wireless
clients to send data simultaneously to avoid packet collision. The 802.11 make use of
the distributed co-ordination function (DCF) to achieve collision avoidance
31
The 802.11 standard was first ratified in 1997 as an extension to the Local
Area Network with a data rate of between 1-2 Mbps. The early version wireless LAN
standard was not welcomed successfully for a long time because of its very low data
rates and conflicting modulation techniques compared to the Ethernet. Therefore was
quickly superseded by the 802.11b in 1999. There are other ratifications that followed
afterwards. Table 2.1 below presents the other 802.11 wireless standard family
specifications in accordance with IEEE and WiFi Alliance.
Table 2.1: 802.11 Wireless Network Standards [3] Standard
802.11
802.11a
802.11b
802.11g
802.11n
Year
1997
July 1999
June 1999
June 2003
Mar 2008
Data Rate
1-2Mbps
54 Mbps
11Mbps
54 Mbps
400 Mbps
Modulation
FHSS / DSSS
OFDM
DSSS
DSSS / OFDM
DSSS / OFDM
Radio Frequency
2.4 GHz
5GHz
2.4GHz
2.4 GHz
2.4 GHz or 5GHz
Channel Width
20 MHz
20 MHz
20 MHz
20 MHz
20 MHz or
40 MHz
The failure to define the 802.11a frequency band components led to late
availability of hardware though it was ratified in the same period as the 802.11b
network standard [4]. The 802.11a has co-existence issues with other 802.11 network
standards as it operates in 5 GHz. This became a major disadvantage in network with
other 802.11 network standards already deployed. As much as the 802.11b network
improved the initial wireless LAN in terms of data rate and reliability, the number of
clients sharing an access point negatively affects its throughput performance for each
client on the 802.11b network. Despite 802.11b standard being widely deployed it is
on the verge of being replaced by 802.11g [4]. The 802.11g has backward device
compatibility with 802.11b. It uses the Orthogonal Frequency Division Multiplexing
(OFDM), which thereby creates a problem for 802.11b devices that cannot sense the
802.11g devices. The throughput for 802.11g network was decreased to 11Mbps in
case of co-existing with the 802.11b.
32
The 802.11g network is prone to much RF interference from microwaves, cordless
phones and other 2.4 GHz devices as compared to 802.11b.
Currently, the IEEE 802.11 Task Group [TGn] is developing a 802.11n
wireless network standard which promises to provide a high data rate compared to
previous ratifications. According to the IEEE Working Group Project Times, its final
ratification is projected to be March 2008 [5]. The 802.11n is expected to use Multiple
Input Multiple Output (MIMO) a new technology that uses multiple transmitters and
receiver antennas to allow increased data throughput. It will also be expected to
operate in both the 2.4 and 5 GHz, thereby being able to support all the 802.11
devices
There are two ways WiFi can be implemented: ad hoc and infrastructure mode.
In the ad hoc mode, the nodes or devices communicate directly with each other
without the use of the access point (AP) whereas in infrastructure mode, the devices
communicate with each other via an Access Point forming a Basic Service Set (BSS).
Ad hoc is also being referred to as “peer to peer “ or Independent Basic Service Set
(IBSS). This ad hoc mode is useful in cases where there are no proper networks. The
infrastructure mode is mostly deployed in companies and institutions.
2.2.2. IEEE 802.15
The 802.15 is a communication specification approved by the IEEE Standard
Association (IEEESA) for wireless personal area networks (WPAN) in early 2001.
The 802.15 standard was designed based on the Bluetooth v1.1 Foundation
Specifications. The IEEE licensed Bluetooth technology from Bluetooth Special
Industry Group (SIG) to adapt and copy a portion of the Bluetooth specifications
based on material from IEEE Standard 802.15.1-2002 [6]. It is a wireless
communication technology and a standard primarily addressing short distance
networking requisite for low-power consumption devices. It communicates the
wireless data and voice among electronic devices based on the low-cost transceiver
microchips. Bluetooth operates in the unlicensed 2.4 GHz providing data rates up to
7kbps. Bluetooth devices cause RF interference problems with the 802.11 standard
33
devices. Most modern Bluetooth standard have a data rate between 2 -3Mbps, such as
the 2.0 + Enhanced Data Rate (EDR) [7].
WPAN could be used for sensor network, which is based on a group of
sensors used to several data reading such as medical measurements of heart rate
(telemetrical environments) [8]. There are some experiments to integrate
communication medium into user’s clothes, as an “intelligent clothes” [9]. Another
example of WPAN useful application is the “in-vehicle” applications, where WPAN
is used to connect the user processing devices such as PDAs to the car electronics [10].
The modern software applications of the fast file download and audio/video
applications require high data rate protocol. Therefore, the high data rate WPAN
MAC protocol has been proposed in 2003 (IEEE 802.15.3). This standard is a
personal wireless protocol that can provide data rate of: 11, 22, 33, 44 and 55 Mbps
[11].
The High Rate (HR) WPAN standard (IEEE 802.15.3) is the next step of the
WPAN group (IEEE 802.15). The first version of WPAN is the IEEE 802.15.1
(Bluetooth) is widely applied in the mobile sets, the laptops and the PDAs. It suffered
from several problems related to the security issues and coexistence with WLAN.
However, security of Bluetooth has been enhanced and coexistence has been
developed but the data rate is still lower than 1Mbps.
Table 2.2 shows a brief comparison between WPAN, WLAN and the
Bluetooth. The HR WPAN standard is proposed to satisfy the QoS and network
requirements of data transmission in low range (about 10 meters) and low power
consumption.
Table 2.2: IEEE 802.15.3 vs. WLAN and Bluetooth [12] 802.15.3 802.11b,g 802.11a Bluetooth1.1
Frequency Band 2.4 GHz 2.4 GHz 5 GHz 2.4 GHz
Data Rate Up to 55 Mbps Up to 22 Mbps Up to 54 Mbps 1 Mbps
Current Drain < 80 mA < 350 mA > 350 mA < 30 mA
Nr. of Video Channels 4 2 5 None
34
Range 10 m 100 m 100 m 10/100 m
A. IEEE 802.15.3
IEEE 802.15.3 standard is designed to enable wireless connectivity of high-
speed, low power, low-cost and multimedia-capable portable consumer electronic
devices [11]-[15]. It provides data rates from 11 to 55 Mbps at distances of greater
than 70 m while maintaining the required quality of service (QoS) for the data streams.
On top of that, this standard is designed to provide simple ad hoc connectivity that
allows the devices to automatically form networks and exchange information without
the direct intervention of the user.
The standard defines the PHY and MAC specifications for HR wireless
connectivity with fixed and portable devices within or entering a personal operating
space. The goal of the standard is to achieve a level of interoperability or coexistence
with other 802.15™ standards. WPAN channel allocation analytical model shows that
channel allocation conflicts occurs in all cases, and is especially severe between IEEE
802.15.3 and IEEE 802.15.4 networks [16]. A personal operating space is a space
around a person or object that typically extends up to 10 m in all directions and
envelops the person whether stationary or in motion.
B. IEEE 802.15.4
In the designing and implementation of wireless sensor network the power
efficiency has a great influence due to its power limitation. The wireless sensor
network applications are required low data rate but it is important the long life of
network. In this context the IEEE 802.15.4 is very suitable protocol for wireless
sensor networks. The IEEE 802.15.4 is low data rate, low power consumption,
flexible, reliable, scalable and extremely low cost communication network which
allows the wireless connectivity in application with limited power and relaxed
throughput requirements.
The main objectives of a LR-WPAN are ease of installation, reliable data
transfer, short-range operation, appropriate level of security and a reasonable battery
35
life. Due to suitability the physical (PHY) and data link layer (MAC) of IEEE
802.15.4 protocol are used by ZigBee specification. Many features of IEEE 802.15.4
are suitable for wireless sensor networks [16]. Assigning a GTS (Guaranteed Time
Slot) to some nodes is also a useful feature for transmitting a critical data in time
sensitive network.
The communication in WSNs depends on which communication protocol is
used. It is considered the latest ZigBee technology. ZigBee is wireless technology that
supports the short-range communication systems applications. In this technology it is
relaxed throughput and latency requirement in wireless personnel area network. It is
the most popular worldwide standard for wireless radio networks in the monitoring
and control fields. The main feature of ZigBee wireless technology is low capacity
and supported by all cheap fixed mobile devices. The main field of application of this
field technology is the implementation of WSNs [17].
ZigBee network supports star, tree and mesh topologies. The physical layout
of star network is just like a star. In the star topology the network is controlled by one
single device known as coordinator. The coordinator is a central node that is linked to
all other devices. The star networks are usually single hop networks. In the tree
networks the coordinator is the root node. The routers are used to form the branches
and end devices are used as leaves nodes. The structure of mesh network is just like
tree but there also some end nodes that are directly connected with each other.
Messages can travel across tree using multiple paths. The root of mesh network is
coordinator node. In mesh and tree topologies the coordinator is responsible for
starting the network and for choosing the certain key parameters but the network may
be extended through the routers. In the tree network routers move the data and control
messages through the network using a hierarchical routing strategy. Tree networks
may be employ beacon-enabled communication as described in the IEEE 802.15.4-
2003 specification. Mesh network allow full peer-to-peer communication. ZigBee
router in mesh network shall not emit the regular IEEE802.15.4-2003 beacons. This
specification describes only the intra PAN networks that is networks in which
communication begin and terminate with in same network.
36
2.2.3. Ultra Wideband (UWB)
Ultra wideband (UWB) communication is a fast emerging technology that
offers new opportunities and is expected to have a major impact on the wireless world
vision of 4G systems [18]. Principal characteristics of UWB signals are their wide
bandwidth (3.1 to 10.6 GHz, and more recently 57-64 GHz) and low intensity, with
levels comparable to that of parasitic emissions observed in typical indoor
environments (FCC part 15: -41.3 dBm/MHz) [19]. UWB radios can have low
complexity and power consumption implying that this technology is suitable for the
mass market deployment of wireless personal area networks (WPAN). Furthermore,
UWB combines high data rates (480Mbps) with localization and tracking features,
and hence introduces many other interesting applications such as safety and homeland
security. For these reasons, UWB is considered a complementary communication
solution within future 4G systems. However, the current high data rate UWB systems
(existing and future evolving multi-gigabit UWB version IEEE802.15.3c at 60 GHz)
are inherently limited to short-ranges of less than 10 m [20].
A significant range extension of UWB communications, through a marriage of
radio and optic channels has been achieved in this thesis. Satisfying conditions such
as integrated laser photodetector system (E/O to O/E conversion and vice versa),
wavelength conversion; up and downstream channel performance, thereby
simplifying customer premises equipment (CPE). As the Wimedia UWB is OFDM
based technique, its impact on any systems’ optical components is prominent and
needs to be addressed for a RoF system. In this thesis the external modulator is
characterised for a range of RF power levels. The wideband (528 MHz per band)
imposes further penalties in terms of second order distortion as it falls within the
transmission band.
2.2.4. IEEE 802.16
WiMAX is the industry name for the set of IEEE 802.16 standards. It is a
point-to-multipoint (PMP) wireless technology that operates in the radio spectrum of
10-66 in the Line of Sight (LOS) and 2-11 GHz in the Non-LOS (WiMAX forum).
WiMAX is an emerging broadband wireless access technology that delivers a carrier-
37
class, high speed at a much lower-cost than the cellular services while providing more
long distances coverage than WiFi. WiMAX was designed to be a cost-effective
technology with a theoretical data rate of 70Mbps over a wide area up to 50km in the
NLOS, at the same time handling high quality voice, data and video services [21].
WiMAX requires LOS for higher frequencies and has compatibility advantage
with other wireless technologies such as asynchronous transfer mode (ATM) and
Internet Protocol (IP). WiMAX is classified as a wireless metropolitan area network
(WMAN). The 802.16 standard was developed after the security failures that
weighted down the progress of IEEE 802.11 wireless networks.
The IEEE 802.16 Working Group in their design for a robust mechanism
incorporated Data over Cable Service Interface Specification (DOCSIS) a solution to
the last mile cable problem. Since security was a major priority in the design of IEEE
802.16 Working Group are busy designing several mechanisms to protect theft of
service and unauthorized information modification and disclosure [22].
2.3 Wireless Networking Characteristic Comparison
There are different wireless technologies that are available today on the
market. The Table 2.3 below provides an overview of these technologies’
specifications.
All the technologies mentioned in the diagram are wireless technologies with
similarities and differences. Since the table above provides an overview of these
wireless technologies, there is need a to engage in a comparative discussion about
their suitability in achieving wireless convergence in providing services to mobile
users and wireless Internet connectivity.
38
Table 2.3: Wireless Technologies Comparison [23]
Wireless Protocol
Data Rates
Radio Spectrum
Airwaves
Range
Bluetooth
1 Mbps
2.45 GHz
Unlicensed
10m
UWB
480 Mbps
3.1-10.7 GHz
Unlicensed
10m
Zigbee
20-250 Kbps
2.4 GHz
Unlicensed
50m
WiFi-a WiFi-b WiFi-g WiFi-n
54 Mbps
11 Mbps
54 Mbps
100 Mbps
5GHz
2.4 GHz
2.4 GHz
2.4 GHz- 5GHz
Unlicensed
30m
100m
100m
30m
EDGE
384 kbps
0.9,1.8,1.9 GHz
Licensed
Several Miles depending on signal free of interference
3G
2 Mbps 1.9-2.1 GHz Licensed Typically 1-5 miles depending on free signal interference
WiMAX
70 Mbps
2-11GHz, 10-66GHz
Licensed and
unlicensed
50 km
2.3.1. WiFi vs. Bluetooth
Bluetooth and WiFi use the same 2.4 GHz unlicensed radio spectrum. These
are important communication technologies that provide different functionalities to
different indoor wireless applications [23]. Bluetooth is a short-range (~10metres)
wireless technology suitable for transferring data file from one device to another in a
close proximity. Bluetooth exist in various devices such as phones, printers, personal
digital assistance, modems and computers.
Due to low bandwidth of Bluetooth, it is not effective to set up a network for
remote applications. Therefore, WiFi technology’s is a better networking
consideration for accessing files compared to Bluetooth. The 16 bit PIN used for
Bluetooth authentication and data encryption is not robust compared to the 80211i
security enhanced protocol used in WiFi [23], [24].
39
2.3.2. WiFi vs. 3G/UMTS
Looking at these two wireless technologies, both facilitate mobility. They
enable mobile devices to be moved around a particular coverage and remain
connected without worry reinstalling cable infrastructure [25]. WiFi mobility is
referred to as local mobility since it is an Ethernet Network extension with a higher
data rate (bandwidth) for a particular entity whereas the 3G offer narrower bandwidth
while covering a large area. In the bid to counter the low data rates offered by 3G, the
telecommunication industry introduced the HSDPA. The HSDPA significantly offers
much better speeds up to 10mbps with lower packet delay. Mobile technologies are
still constrained by limited coverage as signals fade and service diminishes as one
moves away from the city (urban area centres). The cellular connectivity for these
mobile technologies to mobile users moves from HSDPA, through 3G, to GPRS and
finally to EDGE if the cellular network connectivity allows [26] as it moves outwards
from the central business district (CBD). The other important similarity of these two
wireless networks is that they are both access technologies. T3G can be referred to as
an access technology and also as an end-to-end service. The major difference between
mobile technologies (3G, GPRS, EDGE, and HSDPA) and WiFi technology is in the
use of the airwaves frequency bands and the equipment for infrastructure. The mobile
technologies operate on a licensed radio spectrum while the WiFi uses the open
unlicensed 2.4 ISM spectrum.
The use of the unlicensed spectrum has major benefits such as cost of service,
quality of service, congestion management and industry structure. Since the mobile
technologies can be used as an end-to-end service, therefore the integration of WiFi in
the mobile technologies might reduce some of the cost involved in mobile networks
[25]. Though the WiFi technology is still facing security challenges, VPN is
becoming more secure though it adversely affects network performance. The UMTS
provides security to its network users by a smart card device referred to as subscriber
identity module (SIM). Therefore, the SIM will contain all the identification
information, which is used to identify users accessing the services and resources on
the network.
40
2.3.3. WiMAX vs. 3G/UMTS
Though the mobile technologies provide ability for users to use it as an access
technology, its data rates speeds are not able to compete with WiMAX. Theoretically,
the mobile technologies have data speeds from 115 kbps for GPRS, 384 Kbps up to
2Mbps for UMTS and 14.4 Mbps for HSPDA while WiMAX offers high data speeds
of up to 70Mbps for coverage of 50 km. The HSPDA can provide a much faster
theoretical maximum data rate of 14.4 Mbps within a kilometer making it only
suitable for short distance access technology [26]. Similar to other mobile
technologies WiMAX has the same capabilities to transmit data and voice. The choice
of using mobile technologies in transmitting voice is more expensive compared to the
VoIP/WiMAX as to WCDMA/HSDPA.
The HSDPA uses a users’ Subscriber Identification Module (SIM) card for
authentication, while WiMAX supports the strong modern cryptographic algorithms,
which is more robust to protect secret data transmissions [27]. The overall cost for the
WiMAX equipment is much lower as compared to the UMTS. The main advantage of
using the UMTS cellular system is that, its infrastructure for 3G, HSPDA, GPRS and
EDGE is available where there is cellular network coverage while the WiMAX
requires a new infrastructure setup for it to operate. Since UTMS are more easily
available it is more preferred for mobile communication than WIMAX.
2.3.4. WiFi vs. WiMAX
These two wireless technologies have common components in their operations
with a major difference in the communication range. There is a need for many WiFi
access point in order to cover the same distance covered by one WiMAX base station.
Hence it is costly to deploy WiFi for longer distances. WiFi is an access technology
suitable for indoor use due to its short ranges as an extension to LAN technology
while WiMAX was designed for long distance, backhauling and optimized for MAN.
As much as the WiFi has an advantage of providing end user access capabilities, it
can only support a limited number of users (not more than 12 per base station)
whereas the WiMAX base station can support an average of about five hundred users.
WiMAX base station has a scheduling algorithm (First-In First-Out), which allocates
41
a variable channel for each subscriber station to minimize the congestion, and
degrading throughput other than the random queue assignment based on MAC
address in WiFi whereas.
WiMAX uses a licensed and unlicensed spectrum where as the WiFi uses
unlicensed spectrum with a limited channel bandwidth of 20 MHz. Due to the fact
that WiMAX can support high bandwidth, low cost of ownership and provides
backhaul capabilities, it can therefore can be considered as the future broadband
access technology to bridge the ‘digital divide’. WiMAX is more reliable and have
better QoS as it was designed with security as a major priority since the frailness of
WiFi [28]. Despite the similarity in equipment cost, WiMAX technology requires a
costly infrastructure while the WiFi can be easily install using low cost access points.
In conclusion, WiFi has been adopted as an extension for Ethernet some years ago
making it a mature technology compared to WiMAX which is currently license
assignments and infrastructure deployment.
2.3.5. Wireless Convergence Infra
In sections detailed discussion, possible ways have been discovered through
which convergence can be achieved. However, there were factors, which were
highlighted, in each proposed wireless convergence according to its functional
suitability, strengths and weaknesses. Bluetooth, UWB, WiFi and mobile technologies
can be categorized as access technologies, whereas the WiMAX can be classified as
service provider technology.
The capabilities for the WiMAX to connect Internet wirelessly at higher data
rates over a long distance made them suitable technology for backhauling. WiFi
among the access technologies mentioned, offers much higher data rates and cheaper
as (since no expense on service) compared to 3G. Though 3G has a greater range than
WiFi, it operates on the licensed radio spectrum, which makes it expensive due to
monthly charges. As much as the mobile technologies can provide backhauling, it is
usually limited in coverage as one move away from the main base tower. However,
the convergence of WiFi and WiMAX is suitable for rural Internet connectivity.
42
2.4 Handover in Wireless Mobile Networks
In conventional cellular networks and wireless networks handover can be
defined as the mechanism by which an ongoing connection between a mobile host
(MH) and a corresponding terminal or host is transferred from one point of access to
the fixed network to another [31]. When an MH moves away from a BS, the signal
level degrades and it needs to switch communications to another BS. It is very
important in any cellular-based wireless networks because ongoing connection should
be maintained during handover. In cellular voice telephony and mobile data networks,
such points of attachment are referred to as BSs and in wireless networks; they are
called access points (APs). In either case, such a point of attachment serves a
coverage area called a cell. Handover, in the case of cellular telephony, involves the
transfer of a voice call from one BS to another.
In the case of wireless networks, it involves transferring the connection from
one AP to another. In hybrid networks, it will involve the transfer of a connection
from one BS to another, from an AP to another, between a BS and an AP, or vice
versa.
For a voice user, handover results in an audible click interrupting the
conversation for each handover; and because of handover, data users may lose packets
and unnecessary congestion control measures may come into play. Degradation of the
signal level, however, is a random process, and simple decision mechanisms such as
those based on signal strength measurements result in the ping-pong effect. The ping-
pong effect refers to several handovers that occur back and forth between two BSs.
This exerts severe burden on both the user's quality perception and the network load.
In this part of this chapter, a discussion of general handover-related issues, and
handover procedures of representative conventional wireless networks [29]-[31] is
presented.
43
2.4.1. Introduction to Handover
Handover is a process of transferring an active mobile user session from one
Base Station (BS) or Access Point (AP) to another in order to keep the user’s
connection uninterrupted. In the traditional circuit-switched wireless networks such as
GSM, handover is employed mainly for maintaining a mobile user’s telephony voice.
The handover in such a circumstance is motivated by the fact that the coverage
area of a single BS transceiver cannot cover the whole service area. The coverage area
of one or more BS transceivers at a single physical site is referred to as a cell.
In Frequency Division Multiple Access (FDMA) based systems, a cluster is a
group of cells in which frequencies are not reused. Clusters can be repeated with
careful planning to minimise interference among cells using the same frequency so as
to enlarge radio coverage as shown in Figure 2.4. Such a flat compound architecture
can be supplemented by more intelligent radio resource management techniques such
as the macrocell/microcell overlay [32], which consists of large-size macrocells and
small-size microcells for balancing network capacity and network control load
associated with handover.
When a mobile user connection to an AP or BS degrades below an acceptable
threshold, it has to switch the session to a neighbouring cell. If the neighbouring cell
employs the same type of access technology, handover across these cells is often seen
being done horizontally. Horizontal handover refers to handover between base
stations using the same type of network interface. This is common in homogeneous
circuit-switched cellular systems such as GSM and Code Division Multiple Access
(CDMA) networks.
Apart from being implemented in circuit-switched cellular systems, horizontal
handover can be utilised for maintaining the continuity of wireless data services. Such
applications are seen in packet-switched cellular systems such as General Packet
Radio Service (GPRS) and Universal Mobile Telecommunications System (UMTS)
networks.
44
The growing demands for high speed wireless data services drive the
development of new access technologies. Wireless Local Area Networks (WLAN)
such as IEEE 802.11 are able to provide high speed access but with small radio
coverage. For example, IEEE standard 802.11g [33] supports a data rate up to
54Mbps but an outdoor coverage of 140 meters. In contrast, the Third Generation
(3G) cellular systems, e.g. UMTS, can offer much wider coverage through more
complex network architecture.
Figure 2.4: Handover Scenarios [32].
However, the data rates they can offer are not unfavourable for many real-time
applications, which need high bandwidth. Thus, an integration strategy of the two
technologies (e.g. 3G UMTS and WLAN) has been proposed and widely accepted in
the literature [34]-[37] as an economical and feasible solution for providing
ubiquitous access [38]. This integration is expected to result in a heavy demand for
handover between heterogeneous wireless networks, which is recognised as an
important feature of the NG wireless networks. Handover between two networks
based on different access technologies is known as vertical handover. Vertical
handover is further divided into two categories: upward (move-out) vertical handover
and downward (move-in) vertical handover. An upward vertical handover occurs
from a BS/AP with smaller radio coverage to a BS/AP with wider coverage. A
downward vertical handover occurs in the reverse direction to an upward vertical
handover. Apart from handover for radio link quality reasons, vertical handover can
45
be initiated for optimising service quality for wireless data services. In this context,
vertical handover has to deal with the heterogeneities existing in the interconnected
wireless networks.
The objective of supporting various forms of handover on a mobile user is to
provide ubiquitous access across heterogeneous wireless networks and a number of
network operators without manual user intervention [39]. This is supplemented by the
demands for comprehensive and personalised services, stable system performance and
service quality [40]. Seamless handover in the NG wireless networks needs additional
capabilities in network architectures, protocols and control mechanisms, all of which
combine to facilitate the smooth interworking of heterogeneous wireless systems.
Accordingly, the interworking raises a number of research issues, which can generally
classified into two categories: Mobility Engineering and Handover Methodology.
Figure 2.5: Handover Issues in the Multiple Wireless Networks.
Figure 2.5 summarises the major issues in each category. Mobility engineering
provides basic building blocks, which underpin handover functionality. Mobility
engineering provides a common platform for all the mobile users, and comes with
network infrastructure. Integration architectures and mobility protocols lie in this
category. Handover methodology, on the other hand, specifies the way, in which
46
handover should be performed. Unlike network protocols being “hardcoded” for all
the mobile users, handover methodology can be made different for each individual
mobile user for optimised service quality. Handover methodology is comprised of
handover strategies and handover algorithms.
2.4.2. Handover Methodology
A. Handover Strategies
With a well-engineered interworking architecture, handover across
heterogeneous wireless networks can be made possible. Considering current
widespread deployment of cellular networks, it is reasonable to assume that a MH is
within the coverage range of at least one base station at all times. The dimensions of a
base station’s coverage depend upon various factors such as network type,
transmission power and so forth. Therefore, a key issue for both network and mobile
user is to reach a decision as to which network would be selected, and how handover
should be handled when a link transfer is necessary. In an interconnected
heterogeneous wireless infrastructure, the coverage of different wireless networks
may be overlapped in some service areas as illustrated in Figure 2.4.
B. Handover Algorithms
Once a handover control scheme is determined for a specific integrated
network, relevant handover procedure can be executed. Handover procedure in a
heterogeneous environment is more complex than that in a homogeneous wireless
network, which is single technology based. As a rule of thumb, the switching of
underlying access technologies should be kept transparent to higher layer applications
of a mobile user during a handover.
During a handover across heterogeneous wireless networks, once the target
network is determined, the following procedure is defined by the interworking
mechanisms including mobility protocols, authentication methods, integration
architecture and so forth. Handover algorithms are employed to determine the target
network for handover, and make the corresponding decision. This is driven by the
47
introduction of “Always Best Connected” (ABC) concept in mobile service
provisioning [41].
Basically, a handover algorithm deals with two essential tasks in a handover:
network selection and handover triggering. Network selection determines where a
mobile user would be switched to when a handover is necessary. Practically, it is
usually complemented by a separated handover triggering process, which decides
WHEN the switching action to the selected network should be triggered. These two
processes are seen as two consecutive steps in a handover as shown in Figure 2.6.
Figure 2.6: Controlled Handover Algorithm.
2.4.3. Handover Issues
In this subsection general issues related to handover that should be taken into
account in any kinds of wireless networks as long as they involve handover. In this
dissertation handover issues are classified into three parts:
(1) Architectural issues
(2) Handover decision algorithms
48
(3) Handover related resource management.
Architectural issues are those related to the methodology, control, and
software/hardware elements involved in rerouting the connection. Issues related to the
handover decision algorithms are the types of algorithms; metrics used by the
algorithms, and performance evaluation methodologies. Handover-related resource
management deals with the maintenance of quality of service (QoS) during handover.
A. Architectural Issues
The issues are concerned with handover procedures that involve a set of
protocols to notify all the related entities of a particular connection that a handover
has been executed and that the connection has to be redefined as shown in Figure 2.7
[29].
In data networks, the MH is usually registered with a particular point of
attachment. In voice networks, an idle MH would have selected a particular BS that is
serving the cell in which it is located. This is for the purpose of routing incoming data
packets or voice calls appropriately. When the MH moves and executes a handover
from one point of attachment to another, the old serving point of attachment has to be
informed about the change. This is usually called dissociation. The MH will also have
to re-associate itself with the new point of access to the fixed network.
Other network entities involved in routing data packets to the MH or switching
voice calls have to be aware of the handover in order to seamlessly continue the
ongoing connection or call. Depending on whether a new connection is created before
breaking the old one or not, handovers are classified into hard and seamless handovers.
Figure 2.8 illustrates hard handover between the MH and the BSs. A hard handover is
essentially a break before make connection. The link to the prior BS is terminated
before or as the user is transferred to the new cell's BS; the MH is linked to no more
than one BS at any given time. In CDMA, the existence of two simultaneous
connections during handover results in soft handover. The decision mechanism or
handover control may be located in a network entity or in the MH itself. These cases
are called network controlled handover (NCHO) and mobile-controlled handover
(MCHO), respectively. In global system for mobile communications (GSM),
49
information sent by the MH can be employed by the network entity in making the
handover decision. This is called mobile-assisted handover (MAHO). In any case, the
entity that decides on the handover uses some metrics, algorithms, and performance
measures in making the decision.
B. Handover Decision Algorithms
Several algorithms are being employed or investigated to make the correct
decision to handover. Traditional algorithms employ thresholds to compare the values
of metrics from different points of attachment and then decide on when to make the
handover. A variety of metrics have been employed in mobile voice and data
networks to decide on a handover.
Primarily, the received signal strength (RSS) measurements from the serving
point of attachment and neighboring points of attachment are used in most of these
networks. Alternatively or in conjunction, the path loss, carrier-to-interference ratio
(CIR), signal to interference ratio (SIR), bit error rate (BER), block error rate (BLER),
symbol error rate (SER), power budgets, and cell ranking have been employed as
metrics in certain mobile voice and data networks. In order to avoid the ping-pong
effect, additional parameters are employed by the algorithms such as hysteresis
margin, dwell timers, and averaging windows. Additional parameters (when
available) may be employed to make more intelligent decisions.
Some of these parameters also include the distance between the MH and the
point of attachment, the velocity of the MH, and traffic characteristics in the serving
cell. The performance of handover algorithms is determined by their effect on certain
performance measures. Most of the performance measures that have been considered,
such as call blocking probability, handover-blocking probability, delay between
handover request and executions, and call dropping probability, are related to voice
connections. Handover rate (number of handovers per unit of time) is related to the
ping-pong effect, and algorithms are usually designed to minimize the number of
unnecessary handovers. Traditional handover decision algorithms are all based on
received power (P).
50
Figure 2.7: Architectural Issues [31].
Figure 2.8: Hard Handover.
C. Handover-related Resource Management
QoS guarantees during and after handover will become more significant and
challenging in the near future since the current trends in cellular networks are:
1. To reduce cell size to accommodate more MHs that will cause more frequent
handovers.
2. To support not only voice traffic but also data and multimedia traffic such as video.
51
One of the issues is how to control (or reduce) handover drops due to lack of
available bandwidth in the new cell, since MHs should be able to continue their
ongoing sessions. Here, two connection-level QoS parameters are relevant: the
probability (PCB) of blocking new connection requests and the probability (PHD) of
dropping handovers. In ideal case, it would like to avoid handover drops so that
ongoing connections may be preserved as in a QoS-guaranteed wired network.
However, this is impossible in practice due to unpredictable fluctuations in handover
traffic load.
Each cell can reserve fractional bandwidths of its capacity, and this reserved
bandwidth can be used solely for handovers, not for new connection requests. The
problem is then how much of bandwidth in each cell should be reserved for handovers.
Figure 2.9: Decision Algorithms in Handover [31].
This concept of reserving bandwidth for handover was introduced in the mid-
1980s [31]. In this scheme, a portion of bandwidth is permanently reserved in advance
for handovers. Since then intensive research efforts have been carried out for
developing better schemes. Most existing bandwidth reservation schemes for
handover assume that the handover connection arrivals are Poisson, and each
connection requires an identical amount of bandwidth with an exponentially
52
distributed channel holding time in each cell [43]-[45]. But, it is known that the
channel holding time of handed-over connections is not really exponentially
distributed [45].
Figure 2.10: Handover Resource Management [31].
Recently, some schemes attempting to limit PHD to a perspective target value
for multimedia mobile cellular networks have been proposed [47]-[49]. A
probabilistic prediction of user mobility has been proposed in [48] based on the idea
that mobility prediction is synonymous with data compression. From the observation
that a connection originated from a cell follows a specific sequence of cells, rather
than a random sequence of cells, the scheme utilizes character compression technique
to predict future mobility of MHs. In [48], a handover probability at some future time
has been derived using the aggregate history of handovers observed in each cell.
These algorithms depend on the mobility history of users for statistical prediction to
guarantee that PHD is maintained below a prespecified target probability. Thus, they
need a large amount of history data for proper operation. A much simpler scheme has
been proposed in [50]. In this scheme each BS counts the number of handover
successes and failures to adaptively change the reserved bandwidth for handover, and
it does not depend on a large amount of handover history data.
53
This chapter described Radio over Fibre technology, wireless and mobile
network technologies and handover-related management and issues in conventional
mobile wireless networks with a special emphasis on hybrid wireless technologies. In
this thesis, an application-controlled handover scheme investigated in mobile wireless
inter-technology and intra technology [51].
2.5 Conclusions
This chapter has reviewed a number of papers and books related to several
issues of the RoF technology such as the structure, components, signalling and
transmission. This chapter also reviews wireless and mobile network group standards
and their history has been presented briefly. The application controlled handover
development presented in this work is a novel approach.
This thesis presents the concept of applying an intelligent application
controlled handover over heterogeneous wireless access technologies. Some examples
of handover protocols have been overviewed in this work. Converging protocols have
some features that are applied in such as supporting fixed mobile network architecture.
Furthermore, self-organised and scalability are some of the crucial features in hybrid
networks that are applied in the proposed development.
This chapter is crucial for the multiple wireless network handover non-expert
readers in order to understand the enhancement presented by this work to the next
generation networks.
2.6 References
[1] A. J. Cooper, “Fiber/Radio for the provision of cordless/mobile telephony services
in the access network,” Electronic Letters, Vol. 26, No. 24, pp. 2054-2056, Nov.
1990.
[2] IEEE Std. 802.11-1997, Wireless Medium Access Control (MAC) and Physical
Layer (PHY) Specifications for Wireless Local Area Networks (WLANs), 1997.
54
[3] Broadcom Corporation (2006), 802.11n: Next-Generation Wireless LAN
Technology. http://www.broadcom.com/docs/WLAN/802_11n-WP100-R.pdf2,
Dec 2010.
[4] Gast M, “802.11 Wireless Networks: The Definitive Guide” , O’Reilly, 494pages
[5] IEEE (2007), 802.11 working group project timelines, retrieved on 4 March 2007.
http://grouper.ieee.org/groups/802/11/Reports/802.11_Timelines.htm, Nov 2010.
[6] Piscataway, N.J, IEEE Approves IEEE 802.15.1 Standard For Wireless Personal
Area Networks Adapted From The Bluetooth® Specification.
http://standards.ieee.org/announcements/802151app.html, Nov 2010.
[7] Fleishman G, Inside Bluetooth 2.0, retrieved on 29 July 2007.
http://www.macworld.com/news/2005/02/09/bluetooth2/index.php, Dec 2010.
[8] Jovanov, E.; Price, J.; Raskovic, D.; Kavi, K.; Martin, T.; Adhami, R.; “Wireless
personal area networks in telemedical environment” Information Technology
Applications in Biomedicine, 2000. Proceedings. 2000 IEEE EMBS International
Conference, pp. 9-14, November 2000.
[9] M. Gorlick, “Electric Suspenders: A Fabric Power Bus and Data Network for
Wearable Digital Devices,” Proceedings of the Third International Symposium on
Wearable Computers, San Francisco, CA, pp. 114-121, October 1999.
[10] Mahmud, S.M.; Shanker, S.; “In-vehicle secure wireless personal area network
(SWPAN)”, Vehicular Technology, IEEE Transactions, Vol. 55, No. 3, pp. 1051–
1061, May 2006.
[11] James P. K. Gilb, “Wireless Multimedia: A Guide to the 802.15.3 Standard”,
ISBN-10: 0-7381-3668-9, Standard Information Network IEEE Press, 2004.
[12] Karaoguz, J.; “High-rate wireless personal area networks” Communications
Magazine, IEEE, Vol. 39, No. 12, pp: 96–102, December 2001.
[13] IEEE Std. 802.15.3-2003, Wireless Medium Access Control (MAC) and
Physical Layer (PHY) Specifications for High Rate Wireless Personal Area
Networks (WPANs).
[14] Richard C. Braley, Ian C. Gifford, Robert F. Heile; “Wireless personal area
networks: an overview of the IEEE P802.15 working group” ACM SIGMOBILE
Mobile Computing and Communications Review, Vol. 4, No. 1, January 2000.
[15] Ramjee Prasad and Luis Munoz, “WLANs and WPANs towards 4G Wireless”,
ISBN-10. 1-58053-090-7, Artech House Publishers, 2003.
55
[16] Ling-Jyh Chen; Sun, T.; Gerla, M.; “Modeling Channel Conflict Probabilities
between IEEE 802.15 based Wireless Personal Area Networks”, Communications,
2006 IEEE International Conference, Vol. 1, pp. 343–348, June 2006.
[17] I.F. Akyildiz, W.Su, Y. Sankarasubmmaiam, E.Cayirci, “A Survey on Sensor
Networks,” IEEE Comm. Magazine, Vol. 38, No. 4, pp.102-114, August 2002.
[18] I. Oppermann, “The Role of UWB in 4G”, Wireless Personal Communications,
Vol. 29, No. 2, pp. 121-133, April 2004.
[19] Federal Communications Commission,
http://www.fcc.gov/oet/info/rules/part15/part15-9-20-07.pdf , Sep 2010.
[20] Ultra Wideband (UWB) Technologies, White paper, Intel Corporation, 2004,
http://www.usb.org/developers/wusb/docs/Ultra-Wideband.pdf, July 2010.
[21] Westech Communication Inc (2005): Can WiMAX Address Your Application,
White Paper, October 2006
www.wimaxforum.org/technology/downloads/Can_WiMAX_Address_Your_App
lications_final.pdf, Oct 2010.
[22] Johnston D., Walker, J. (2004). Overview of IEEE802.16 Security,
http://mia.ece.uic.edu/~papers/WWW/Bubbles/segment/WiMax_Security.pdf
May 2010.
[23] Groupe Ingenico, “WiFi vs. Bluetooth Two Outstanding Radio Technologies
for Dedicated Payment Application”,
http://www.ingenico.com/INGENICO_GALLERY_CONTENT/Documents/corpo
rate/whitepaper/Whitepaper-Wi-fi-vs-Bluetooth.pdf (Sep 2010).
[24] Crookston R (2004): “Bluetooth vs. WiFi”,
http://www.verifonedevnet.com/VeriFone/Attachment/20040804/RetailTechnolog
_407_28_316.pdf, July 2010.
[25] Lehra W and McKnight L.W, “Wireless Internet access: 3G vs. WiFi? “,
Telecommunications Policy 27, pp. 351–370, 2004.
[26] Vidalis S, “A Critical Discussion of Risk and Threat Analysis Methods and
Methodologies: WiFi Hotspots vs. 3G/HSDPA vs. WiMAX”, Dec 2004.
[27] Lurie S, “HSDPA vs. WiMAX: Comparing Characteristics and Prospects of
Datacom Technologies”, http://www.digit-life.com/articles2/mobile/wimax.html
Oct 2006.
56
[28] Mylavarapu R, “Security Considerations for WiMAX-based Converged
network”, http://rfdesign.com/mag/508RFDF1.pdf. Dec 2005.
[29] G. P. Pollini, “Trends in Handover Design,” IEEE Communications Magazine,
Vol. 34, No. 3, pp. 82-90, March 1996.
[30] N. D. Tripathi, J. H. Reed, and H. F. Vanlandingham, “Handoff in Cellular
Systems,” IEEE Personal Communications,Vol. 5, pp. 26-37, December 1998.
[31] K. Pahlavan, P. Krishnamurthy, A. Hatami, M. Ylianttila, J. Makela, R. Pichna,
and J. Vallstr¨om, “Handoff in Hybrid Mobile Data Networks,” IEEE Personal
Communications, Vol. 7, No. 2, pp. 34-47, April 2005.
[32] N. D. Tripathi, J. H. Reed, and H. F. Van Landinoham, "Handoff in cellular
systems," Personal Communications, IEEE, Vol. 5, pp. 26, 1998.
[33] IEEE, "Part 11: Wireless LAN Medium Access Control (MAC) and Physical
Layer (PHY) specifications - Amendment 4: Further Higher-Speed Physical Layer
Extension in the 2.4 GHz Band," in IEEE Std. 802.11g, IEEE, June 2003.
[34] K. Ahmavaara, H. Haverinen, and R. Pichna, "Interworking architecture
between 3GPP and WLAN systems," Communications Magazine, IEEE, Vol. 41,
pp. 74, 2003.
[35] A. K. Salkintzis, "Interworking techniques and architectures for WLAN/3G
integration toward 4G mobile data networks," Wireless Communications, IEEE,
Vol. 11, pp. 50, 2004.
[36] 3GPP, "3GPP system to Wireless Local Area Network (WLAN) interworking;
System description (Release 6)," in TS 23.234: 3rd Generation Partnership Project,
2005.
[37] I. F. Akyildiz, S. Mohanty, and X. Jiang, "A ubiquitous mobile
communication architecture for next-generation heterogeneous wireless systems,"
Communications Magazine, IEEE, Vol. 43, pp. 29, 2005.
[38] Y.-F. R. Chen and C. Petrie, "Ubiquitous Mobile Computing," IEEE Internet
Computing, Vol. 7, pp. 16 - 17, April 2003.
[39] C. Politis, T. Oda, S. Dixit, A. Schieder, H. Y. Lach, M. I. Smirnov, S. Uskela,
and R. Tafazolli, "Cooperative networks for the future wireless world,"
Communications Magazine, IEEE, Vol. 42, pp. 70, 2004.
57
[40] C. Kalmanek, J. Murray, C. Rice, B. A. G. B. Gessel, R. A. K. R. Kabre, and
A. A. M. A. Moskal, "A network-based architecture for seamless mobility
services," Communications Magazine, IEEE, Vol. 44, pp. 103-109, 2006.
[41] F. Zhu and J. McNair, "Multiservice Vertical Handoff Decision Algorithms,"
EURASIP Journal on Wireless Communications and Networking, Vol. 2006, No.
2, pp. 13, May 2006.
[42] D. Hong and S. S. Rappaport, “Traffic model and performance analysis for
cellular mobile radio telephone system with prioritized and non-prioritized
handoff procedure,” IEEE Transaction on Vehicular Technology,” Vol. 35, No. 3,
pp. 77-92, 1986.
[43] E. Del Re, R. Fantacci, and G. Giambene, “Handover and Dynamic Channel
Allocation Techniques in Mobile Cellular Networks,” IEEE Transaction on
Vehicular Technology, Vol. 44, No. 2, pp. 229-237, May 1995.
[44] R. Ramjee, R. Nagarajan and D. Towsley, “On Optimal Call Admission
Control in Cellular Networks,” in Proceedings IEEE Infocom 2000, Vol. 1, pp.
43-50, Mar. 2003.
[45] X. Luo, I. Thng, and W. Zhuang, “A Dynamic Channel Pre-reservation
Scheme for Handoffs with GoS Guarantee in Mobile Networks,” in Proc. IEEE
International Symposium on Computers and Communications, pp. 404-408, 2002.
[46] C. Jedrzycki and V. C.M. Leung, “Probability Distribution of Channel
Holding Time in Cellular Telephony Systems,” in Proceedings IEEE Vehicular
Technology. Atlanta, GA, Vol. 1, pp. 247-251, 2006.
[47] F. Yu and V. C.M. Leung, “Mobility-Based Predictive Call Admission
Control and Bandwidth Reservation in Wireless Cellular Networks,” Proc. IEEE
INFOCOM 2001, pp. 518-526, 2001.
[48] S. Choi and K. G. Shin, “Adaptive Bandwidth Reservation and Admission
Control in QoS-Sensitive Cellular Networks,” IEEE Transaction on Parallel
Distributed. Systems, Vol. 13, No. 9, pp. 882-897, Sep. 2002.
[49] IEEE Recommended Practice for Multi-Vendor Access Point Interoperability
via an Inter-Access Point Protocol Across Distribution Systems Supporting IEEE
802.11 Operation. IEEE Draft 802.11f/D5, Jan. 2003.
58
[50] H. B. Kim, “An Adaptive Bandwidth Reservation Scheme for Multimedia
Mobile Cellular Networks”, Proceedings of IEEE ICC 2005, May. 2005, Seoul,
Korea.
[51] S R Chaudhry & H S Al-Raweshidy “An Application Controlled Handover for
Heterogeneous Radio in Wired Wireless Integrated Networks”, Wireless Personal
Multimedia Communications IEEE (IST 07), Budapest, Hungary, July 2007.
59
Chapter 3
Application Controlled Handover
Proposed Architecture __________________________________________________________
3.1 Introduction
The Wired Wireless Integration Network (WWIN) can be categorised as Fixed
Mobile Convergence (FMC). FMC means the convergence of the existing Wired
Network and Wireless Network. Therefore a mobile device needs the function of
connection and control to the FMC infrastructure. An application controlled handover
is developed in this chapter, which keeps channel continuity in the wired wireless
synergy network environment that consists of 3G (UMTS) + WLAN + WPAN
(UWB) and optical fibre network. This chapter presents a novel handover mechanism
that transmits and receives data by using the proposed application selection criteria. It
maintains the channel and the seamless transmission from mobile device to the remote
optical fibre network, to provide real time service continuity for multimedia traffic.
The integration of all modes of electrical communication is under great
progress and the fact remains that there is an undeniable and healthy competition
between wire and wireless networks [1]. With the rapid development of
communication and network technologies, traditional voice networks and data
networks have been gradually merged together into multimedia networks which
supply various services ranging from email, web browsing to telephony, visual
conference and video on demand. The horizon of the integrated multimedia network
is extended further to incorporate wired, wireless and cellular networks.
The progress of beyond 3rd generation or 4th generation (B3G/4G) networks
will depend largely on the close coordination between mobility, resource and quality
of service (QoS) management schemes [1]. The B3G/4G networks are expected to
serve stationary as well as mobile subscribers under dynamic network conditions and
60
provide any type of service – anytime - anywhere and anyhow. Here a discussion of a
hierarchical, layered and modular B3G/4G network architecture with wired-wireless
integrated network coordination and distributed network functionalities has been
presented.
An Application Controlled mechanism for control/signalling for multi-radio is
adopted to facilitate enhanced wireless system performance. Reconfigurable
architecture for multi-application, multi-homed and multi-service supported mobile
device has been presented to enable seamless mobility across different access
technologies including wired and wireless infrastructures.
In an integrated 3G+WLAN+WPAN, as illustrated in Figure 3.1, more data
and multimedia applications are carried end-to-end over the current Internet protocol
infrastructure. Fundamentally, Wired Wireless Integrated Networks (WWIN) will
reshape the telecommunications industry. Technological changes in the
telecommunication industry have not only led to the creation of new trends in the
deployment of next generation networks, but also to the convergence of
telecommunication technology sector. In particular, the so-called WWIN has become
a reality due to the market demand and industrial competition.
FMC enables users to work with existing wired and wireless networks. UMTS
has nationwide coverage and access to UMTS network is achieved by using cellular
networks. WLAN and UWB are free of charge communication air-interfaces. WLAN,
UWB and UMTS can serve the Internet core network on the Transmission Control
Protocol/Internet Protocol (TCP/IP). TCP/IP enables a user to adopt client server
communication by accessing an individual network.
61
Figure 3.1: An Integrated Wired Wireless Network Environment.
This chapter is organised as follows. Section 3.2 highlights the research
contribution and related work done in this area. Section 3.3 presents the current
WWIN capabilities and the research challenges. Section 3.4 presents the proposed
WWIN architecture and organisation of communication environment for FMC. In
Section 3.5, the design of ACH has been discussed in detail. Section 3.6 demonstrates
simulation network model, scenarios and layout for proposed technique. Section 3.7
62
shows the results leading to the conclusions of this chapter that are presented in
section 3.8.
3.2 Global Contribution and Related Work
The architecture classification for WWIN is defined as: Intracellular, Inter-
cellular and Extra-cellular. In Intracellular architectures, each Access Point or Base
Station (AP/BS) communicates with any mobile node (MN) in its coverage area in a
multi-radio mode.
Figure 3.2: Hierarchical Classification of Wired Wireless Network Architectures.
On the other hand, in Inter-cellular architectures, each AP/BS communicates
with any MN in the coverage area of other APs/BSs in a multi-radio mode. Finally,
Extra-cellular architectures enable an AP/BS to communicate with MNs that are not
in the coverage area of any AP/BS. Any of these classifications may employ
dedicated radio stations. Furthermore, Intra/Inter cellular ad hoc mode architectures
can be classified as multi-radio systems in which MNs act either in single-radio mode
63
or multi-radio mode. Figure 3.2 illustrates this classification with respect to the well-
known references such as [2]-[12] in this research era.
3.3 Current WWIN Capabilities
In the future Internet where wired and wireless networks are integrated, traffic
volume of wireless user should expand greatly and QoS of wireless user will be one
of the most important technical issues. In wireless network environment TCP, one of
most important protocol of TCP/IP, suffers from significant throughput degradation
due to the lossy characteristics of a wireless link. Therefore, in order to design the
next generation networks, it is necessary to know how much improvement can
wired/wireless integrated network brings.
Throughput of TCP connection is well known to be dependent upon both
packet loss and application response delay. Several papers have been published in
wireless TCP, for example, [13] evaluates throughput performance of a single
wireless TCP connection experimentally and [14] analyses the throughput
performance mathematically using fluid flow model. These papers only dealt with
wireless TCP and did not take care of interaction of wireless and wired TCP in FMC
environment.
3.3.1. Wired Wireless Convergence Era
Wireless broadband is the next phase of the wireless evolution. While 3G has
already taken off, the R&D effort in the wireless area is concentrating on higher bit
rates, high bandwidth availability, less power consumption, high security and etc. The
pace of introduction of new radios and architectures continue to increase; therefore,
the traditional era based on long standardization processes is not possible anymore.
The broadband battlefield may change the classical operator model due to new
entrants using alternative technologies in other frequency ranges for different usage
scenarios under different regulatory conditions. Additionally, new proprietary systems
are being proposed for broadband, one of which is convergence of wireless and
optical fibre.
64
Optical fibre media will do better transmitting high frequency radio signals
over a longer distance because optical fibre cables are immune to EMI, provide
excessive bandwidth in the region of THz, and are more secure. More importantly, its
ability to handle very large numbers of radio frequency (RF) and digital data signals
at different frequencies simultaneously over one optical cable with less attenuation
which makes it a better choice over conventional copper based wiring achieving
higher data rates and better Quality of Service (QoS) in terms of bandwidth hungry
and real time application. In the network side, the generic solution to seamless
connectivity lies in the IP protocol that runs over a range of heterogeneous transport
physical mediums Fixed and wireless access technologies are further being integrated
at the service layer by devising new control plane solutions such as Unlicensed
Mobile Access (UMA) and IP Multimedia System (IMS). In addition, the industry is
moving rapidly towards integrating all types of electronics, including entertainment
systems, and this has generated a lot of interest in convergence of wireless and fixed
approach.
3.3.2. The Research Challenge
Original work in this chapter is the development and realisation of handover
technique for multiple access technologies, which can provide data transmission and
service continuity in the FMC environment The algorithm uses a cross layer solution
to operate on transport layer between wired and wireless network infrastructure. The
PDA, Mobile phone and a Laptop were used in the wired and wireless network
simulation environment. The devices have installed handover algorithm and moves
freely in the WWIN environment.
The handover provides service continuity and data transmission by selecting
appropriate wireless technology depending on the required data rate for specific
application. The traffic was recorded at the remote server that was located in wired
network and traffic was transported from devices that were located in wireless
network.
65
3.4 Proposed Modification in WWIN Architecture
The existing integrated high-speed wired and wireless networks are shown in
Figure 3.3, it also reveals general constitution of FMC. Both Wired and Wireless
networks have separate service platform, as shown in Figure 3.3. FMC is managed by
using Integration Service Server (ISS) [15]. FMC network supports data service in the
existing voice/data based wireless infra, and the existing high bandwidth wired
network infrastructure.
Figure 3.3: Proposed FMC Architecture for Wired Wireless Integrated Network.
3.4.1. Constitution of Communication Environment for FMC
The proposed FMC environment is based on multi-radio such as UMTS,
WLAN, UWB and optical fibre network. Proposed constitution of communication
environment for FMC is shown in Figure 3.1. This research work realizes FMC
environment by using TCP/IP and transport layer between wired wireless networks
not by using ISS. It can be seen from Figure 3.1, there is no physical ISS for proposed
FMC. Application controlled handover uses transport layer and TCP/IP for each
network constituting FMC. So accordingly there is no physical ISS. In the architecture,
mobile devices select radio access technology according to the user requested
application. Light FTP data can be accessed using UMTS cellular network, heavy
66
FTP and Database applications (remote account management and payroll systems for
a multinational corporate) may use WLAN and for video conferencing UWB is
prominent candidate supporting up to 200 Mbps [16].
Figure 3.1 shows a user carrying mobile devices in the wireless network
domain having access to UMTS, WLAN and UWB air interfaces. Mobile device
network protocol stack can communicate cross layer PHY to APP Layer to detect the
transmission of data, handovers to the suitable air-interface and sends it to the remote
server. Once server gets the request, it records real time data transmitted from user’s
device.
3.4.2. Application Controlled Handover Classification
The definition of Application Controlled Handover is explained as follow.
When the mobile device moves in wireless network domain having access to UMTS,
WLAN and UWB, a certain process is required in order to switch and maintain
network connection channel. In the proposed handover technique, if average required
data rate for a specific application is greater than the threshold of achievable data rate,
a handover to the acquired radio access initiates. The handover is classified at the
point of various communication resources as shown in Table 3.1.
Table 3.1: Classification of Handover At the point of domain controlling network
• Inter domain handover: The action of handover in different domain networks. It is called
roaming.
• Intra domain handover: The action of handover in same domain networks.
At the point of an object controlling handover
• Controlled Handover: The device checks the signal strength and required data rate and
controlling manager decides to trigger handover.
At the point of an access platform
• Vertical Handover: The action of handover in different platform environment.
• Horizontal Handover: The action of handover in same platform environment
At the point of handover service
• Connectionless: During handover, the home server keep the position information and route of
67
device
• Switching Platform: Handover depends on the required bandwidth to perform a particular
task; i.e. for heavy real time application such as video conferencing; a handover from WLAN to
UWB due to high bit rate requirements.
In this chapter, the proposed FMC architecture uses vertical handover for
application controlled mechanism.
3.5 Information Element of the Proposed ACH
An ACH has been proposed which keeps the continuous channel service for
wired and wireless networks. The mobile devices not only sense the signal strength
but also determine the application type and initialise the handover process. A soft
handover has been considered in the simulation network due to the session handling
for connections to more than two radios e.g. connections to UMTS, WLAN and UWB
can be maintained by one device at the same time. It is a trade off between the one
channel occupancy (hard handover) and the QoS in terms of soft handover’s make-
before-break (session continuity).
This section elaborates the application controlled based network protocol
stack, the design of network components and finally the soft handoff process model
for ACH.
3.5.1. Application Controlled Network Protocol Stack
Figure 3.4 depicts the protocol stack for the proposed handover technique.
Seamless bridging is supported natively by this protocol, which enables both
information and process data to be easily exchanged in heterogeneous environment.
Device handover manager adopts the lower layers of the protocol stack (PHY).
Despite the relatively low speed of UMTS (up to 384 Kbps), it is able to guarantee
deterministic behaviour. As long as the load on the network is well below the
theoretical available bandwidth, handover technique guarantees that any low or high
data is delivered to the server within the suitable application response time.
68
Figure 3.4: Application Controlled Network Protocol Stack.
3.5.2. Selection of Random Application by Probability Function
Real-life devices run several applications (such as word processing and
Internet browsing) simultaneously (“user profile” as defined in OPNETTM) [17]. The
user profile (Students, Engineers, Researchers, etc.) is identified depending upon the
user specialisation, professional background and interests. The user profile is a set of
applications, which lead to a particular level of power consumption. However,
prediction of next application “consumption weight” (heavy or light) and duration is
not an easily solved problem. Therefore, this work has applied the probability theory
(Probability Distribution Function PDF) in order to simulate the running applications
on the devices [18]. Battery analytical models have been presented in [19]-[23].
Although the literature has proposed interesting battery lifetime models, these
models are not easily implemented because they focus on physical study of power
source and the wireless device [19]-[23]. The work proposed in this chapter presented
69
a new approach based on PDF, which leads to simplicity in implementation.
Applications in the wireless devices (such as computer games and word processing)
are manually turned ON and OFF by the user. Therefore the probability function
could be used to provide an accurate estimation for the expected application type (i.e.
heavy or light power consuming).
Accordingly, power consumption and battery lifetime could be calculated and
provided to the simulation model of the device for further analysis and decision such
as route selection or switching the device ON/OFF. The proposed algorithms for
battery power model and charger (shown in Figure 3.5 and Figure 3.6) are quite
simple.
Figure 3.5: Battery Power Management Flowchart.
70
Figure 3.6: Charger Model Flowchart.
These models are implemented by C/C++ using network simulation tool such
as OPNET and NS2. The PDF Equation (3.1) is applied to identify the probability of
power consumption due to the instinctual application [24].
P(X ≤ x) = P(x)dxxmin
x
∫ (3.1)
In Equation (3.1), is the probability when X is less or equal to a certain
value x. The simulation model proposed in this work could be applied for simulated
wireless devices in any air-interface such as WLAN, WPAN, UMTS and WiMAX.
3.5.3. Battery Lifetime Model of the Wireless Device
The goal of the simulation model is to demonstrate the battery lifetime for
several wireless devices and charger process algorithm. The battery level capacity (the
device lifetime) is one of important criterion, which could be used for node selection
in route discovery in ad hoc networks or PNC selection in WPAN. The device battery
71
power capacity is modelled using the rules shown in Equation 3.2, where Pch arg e is the
probability of charging a battery and is the Remaining Battery Power in
Percentage. All of the values used in Equation 3.2 are assumed in the proposed model.
Pch arg e =
25% if Batpower ≥ 75%33% if 40% < Batpower < 75%50% if Batpower ≤ 40%
⎛
⎝
⎜⎜⎜⎜
⎞
⎠
⎟⎟⎟⎟
(3.2)
Equation 3.2 is proposed based on the assumption of a logical user behaviour
of charger usage. In this aspect, the probability of using a charger in the battery-
operated device is a function of the battery remaining power. In a typical regular-user
behaviour, the probability of charging a battery is dependent on the remaining battery
power of the device. The device is most probably charged when the remaining power
is low. On the other hand, if a device has a larger remaining battery power (defined by
), the probability of charging the device battery ( ) is lower. If the
battery remaining power is lower, the probability of charging the battery is increased.
It is assumed that if the battery remaining power is higher or equal to 75% of the
maximum value, the probability of charging the battery is 25% as shown in Equation
3.2. If the battery remaining power is equal or less than 40%, the probability of
charging the battery is 50%. Typically, the lower the battery level is, the higher the
probability of charging.
Power Unit management algorithm is shown in Figure 3.5, where the battery
value is updated every 30 seconds. The charger model is shown in Figure 3.6. If the
charger indicator is not present (the charger is not connected), the battery power level
of the device will decrease due to power consumption in the device. Power
consumption could be estimated according to the application in the device. Assuming
that the power consumption is divided into four levels (Idle, Low, Medium and High),
the algorithm presented in Figure 3.5 is responsible for random switching between
these levels.
The device applications have been divided into four categories as outlined in
Table 3.2. The lowest power consumption rate is shown when the device is idle. At
72
this state, only the device essential parts such as the display and the memory are
activated. The processor load is at minimum required level in order to maintain the
operation system running. The values in the fourth column (in Table 3.2) are
calculated according to the assumption made in this work. Based on this assumption,
if the device is idle, maximum battery life is reached (250 minutes).
However, if a low-power consumption application is used (power
consumption rate is 0.6 of maximum value per minute), the device would run for 167
minutes. The selection process for the battery consumption rate uses a uniform
distribution to switch between the four values. The selected power consumption value
will be subtracted from the current battery-power level. However, if the charger
indicator is ON, the device battery will recharge at a rate of 2%/min.
Table 3.2: Power Consumption Rates in Battery-operated Devices
# Applications Consumption Rate (Max. Value per
min) Max Battery
Lifetime (min) 1 Idle (No Application Running) 0.4 250 2 Low Power Consumption App. + Idle 0.6 167 3 Med. Power Consumption App. + Idle 0.8 125 4 High Power Consumption App. + Idle 1.0 99
This value will be added onto the current battery level up to value 100 (full-
battery) as shown in Figure 3.5. If the probability density function presents that the
charger is connected, the power unit model will go to the Charger Indicator state
‘Yes’ while ‘No’ indicates that a charger is not available. The presence of a charger is
calculated every 15 minutes. Hence, the Power Unit management model will maintain
the current Charger Model indicator state for 15 minutes during its repeated 30
seconds iteration.
3.5.4. Battery Lifetime Simulation Results
As shown in Figure 3.7, simulation results in OPNETTM have been obtained
using five wireless devices (DEV1, DEV2, DEV3, DEV4 and DEV5). The simulation
time is five hours. The simulation results show that the random behaviour of charging
73
and battery power consumption due to the four categories of the applications as
shown in Table 3.2
In the simulation results (Figure 3.7), the positive slop (rising curve) is the
time when the charger is connected. In this simulation, devices are assumed identical.
Therefore the charging curves are at the same rising slop. On the other hand, the
“negative” slop of the remaining battery power reflects the power consumption
(battery discharging) at that time.
Figure 3.7: Battery Power in Mobile Devices.
For example, in the period between 32 minutes and 120 minutes, Dev_4 and
Dev_5 illustrate three rates of power consumption pointed as “A”, “B” and “C”.
Period “A” is for lower power consumption when the device is at the idle mode where
the power consumption source is the device physical parts such as RAM and monitor.
The power consumption of period “A” is referred to as number one in Table
3.2. Power consumption at “B” represents light applications such as word processing
tasks. This power consumption is categorised as number two in Table 3.2. Finally,
74
number three in the power consumption levels in Table 3.2 is shown in the period “C”
(Figure 3.7).
As illustrated in Figure 3.7, the power consumption rate is randomly changed
between the four levels, which are presented in Table 3.2. These applications simulate
the real-life user profile, where the devices switch between multiple tasks according
to the user requirements.
At the same time, the charger usage in the simulation results follows Equation
3.2. Several applications are shown in Figure 3.7 distributed on random basis. This
simple simulation model could be applied in any network simulation tool in order to
utilise the device lifetime in the route selection or devices comparison.
3.5.5. Application Controlled Handover Network Components
The design of components for mobile device is illustrated in Figure 3.8. It
consists of the (I) data component that can exchange data with external device, (II) the
signal acquisition component that can receive radio signal strength of wireless
network, (III) socket control component that controls the communication resources
inside the device and the application control handovers to the wireless network
platform according to the demand of quality data transmission requirements.
Application controlled handover used a virtual socket to connect UMTS,
WLAN and UWB network access infra. It is designed as if there is only one virtual
socket device [25].
However UMTS, WLAN and UWB can access the data by using data buffer in
the virtual socket and control of access is designed and serviced by application
controlled handover technique. The application handover manger needs to decide the
connectivity between the mobile devices and AP/BS [26]. In addition, wireless radio
selection criteria were mainly emphasised.
75
Figure 3.8: The Network Components of ACH.
It is shown in the Figure3.9, signal control component can control all the three
signals together between data control and socket control component.
The Signal acquisition component provides the signal strength of AP/BS.
Application controller is connected to application manager for handover process.
Figure 3.9 shows the implementation of how the components of Figure 3.8 are
connected with the TCP. Data control component is connected with external device
and it collects data from external device (server) periodically.
76
Figure 3.9: Data and Socket Control Components of ACH.
3.5.6. Application Controlled Handover Process Model
Soft handover sets up a new channel on condition that it maintains the existing
channel where as hard handover sets up a new channel after breaking the existing
channel. The socket control component controls a virtual UMTS, WLAN and UWB
socket in Application Controlled Handover.
A socket control component controls three sockets; UMTS, WLAN and UWB
respectively. The buffer is connected to hidden UMTS, WLAN and UWB sockets. A
soft handover process design is shown in Figure 3.10.
77
Figure 3.10: ACH Process Model Flow.
3.6 Proposed Network Model Layout
Network model implementation has been divided into two major scenarios
(I) Multiple radios over fibre network without ACH (base network model)
(II) ACH based multiple radios over fibre network.
78
Both scenarios are simulated parallel and compared on the basis of same
parameters; the details are in the following sub sections.
The client subnet in Figure 3.11 represents the wireless client with multi radio
mobile devices capable of WLAN, UWB and UMTS connectivity in OPNETTM [17].
The AP supports both PHY layers of WLAN and UWB. UWB model used here is 10-
12 meters in range with a data rate up to 200 Mbps. Network was simulated to analyse
the scope of the simulation and to achieve reasonable simulation time.
Figure 3.11: Network Simulation Model Layout.
3.6.1. Multiple Radios over Fibre Network
In this scenario, the behaviour of three wireless technologies UMTS, WLAN
and UWB were examined within the framework of a deployed FMC to simulate the
WWIN without an efficient handover technique. Firstly, the mobile device
79
communicates to remote server using simple http application with UMTS
infrastructure. The data traffic changed from HTTP to FTP (Databases) and then to
Video conferencing respectively. The WLAN and UWB were accessed via AP, which
was connected to wired backbone server through a central hub using optical fibre (2
Gbps) simulating a real life office environment [27], [28]. An IP gateway (i.e. an
enterprise router) connected the wireless network to an IP cloud used here to represent
the backbone Internet connectivity. The remote server supports three kinds of traffic
FMC such as HTTP, FTP/Heavy FTP (Databases) and real time video
calling/conferencing.
3.6.2. Multiple Radios over Fibre Network with ACH
In this second entire network simulation settings and configurations were kept
same as for the first scenario. The only addition is mobile devices have been equipped
and updated with ACH functionality.
3.6.3. Network Parameters
ACH made the data exchange very bandwidth efficient and enhance system
performance as discussed and presented in this section. The simulation network
parameters and traffic type for both scenarios are described in Table 3.3 and Table 3.4
respectively.
Table 3.3: Network Simulation Parameters
Number of Nodes
3
Simulation Area
500x500 Meters
Simulation Time
60 Min
Multimode Fibre
Channel bit rate
2Gbps
80
Table 3.4: Network Data Traffic Profiles
3.7 Performance Evaluation and Discussions
In this chapter, an imperative performance evaluation has been carried out for
multiple radios over fibre network with a special focus on the performance of the
UMTS, WLAN and UWB when subjected to rich multimedia applications. Real time
traffic was introduced in the network in the form of video conferencing. Video
conferencing encompasses data, voice and images and adequately represents the
ultimate heavy multimedia traffic.
No extensive network management techniques were configured but the
application controlled handover technique has been used to identify the effect of
multiple wireless radio interfaces on the performance of the optical fibre network.
The performance measurements those are relevant to network traffic, such as
wireless and cellular link utilisation in terms of throughput was obtained and
discussed. More specifically, the response times of applications (HTTP, Database and
Video Conferencing) at the mobile devices were also compared to ascertain the user
alleged QoS.
Additionally some critical network parameters such as media access delay,
end-to-end packet drop, bit error rate and network power consumption has also been
evaluated for both scenarios. A simulation time of one hour was set for both scenarios
to achieve feasible results.
APPLICATIONS
H= Heavy L=Light
UWB / WLAN / UMTS
devices
HTTP
FTP
Real Time-Video Conf.
No Handover
L
H
H
App Controlled Handover
L
H
H
81
3.7.1. Network Throughput
Comparing the multiple radios over fibre network in both scenarios showed
that a high network throughput has been achieved for the Application Controlled
scenario. Figure 3.12 presents a comparison of average network throughput for same
type data traffic run for both scenarios.
One scenario used ACH mechanism and results in 83% improvement on the
throughput of the network. Application controlled optical network has reached up to
an average of 12000 bps as compared to 2000 bps by the network without handover
technique.
Figure 3.12 Network Average Throughput.
3.7.2. HTTP Client Server Response Time
There was enormous improvement in the HTTP response time of the optical
network with ACH when subjected to http traffic as shown in Figure 3.13.
The result shows that optical network with handover can handle higher load
than the base model (without handover) due to the selection of appropriate wireless
air interface. It can be seen clearly from Figure 3.13 that both network models
performed similar when the network data traffic was kept simple in the start of
82
simulation time. Whereas, after introducing heavy http browsing, the response time
reached to the 27.5 seconds for optical network without handover technique. Whereas
optical network with handover performed excellent comparatively and reached an
average response time of 4.5 seconds which is 23 seconds less than the base model.
Figure 3.13: HTTP Average Response Time.
3.7.3. Database Client Server Response Time
A similar improved performance was seen in database response time (FTP)
during the simulation. Database needs higher authentication and security information
comparing to simple HTTP application. Therefore it enhanced the load on the
communication technology.
After 3000 seconds of simulation time, the database response time reached up
to 1000 seconds for base model while using UMTS for database access as shown in
Figure 3.14. However, application controlled handover improved network response
time by switching to WLAN. Database response time for controlled handover reached
to 10 seconds, which is 100 times less than the base network.
83
Figure 3.14: FTP Average Response Time.
3.7.4. Video Conferencing Response Time
Optical fibre network without handover used UMTS service for real time
video conferencing, but heavy multimedia video data transmission overloaded the
radio link. The Handover technique in optical network enhanced the response time
from the video server. An average improvement of 3 packet/sec for video
transmission delay can be seen in Figure 3.15.
Figure 3.15: Video Conferencing Average Response Time.
84
3.7.5. Network Multiple Media Access Delay
In Figures (3.12-3.15), it was seen that ACH improved the network efficiency
and enhanced the performance enormously. ACH works on cross APP–TRA-PHY
layers; the algorithm’s working architecture is based on the selection of particular
wireless technology for specific application type. Handover mechanism increased the
medium access delay of the network because of handover between different wireless
access technologies, whereas with base network it resulted in 0.00005 seconds less
delay as shown in Figure 3.16.
Figure 3.16: Network Average Media Access Delay.
The medium access delay is function of handover algorithm and traffic
characteristics. That is why, where this algorithm enhanced the network efficiency, it
led to increase in medium access delay too. Once the network was established a linear
behaviour was observed from both network scenarios.
3.7.6. Network Packet Drop
Figure 3.17 shows the peak-to-peak network average packet drop comparison
for both network scenarios. Base network shows very non-linear behaviour because of
high variation in transmission. It has dropped up to 18 packets/second and gradually
improved to an average of 10 packets/second. On the other hand, application
85
controlled handover performed well by reducing the packet drop up to 83%
comparatively.
Figure 3.17: Network Average Packet Drop.
3.7.7. Network BER and Power Consumption
Base network model showed very non-linear behaviour for both BER and
power transmitted. The glitches in the graphs are because of the variation in data
traffic and high bit rate demand. It can be seen that ACH have reduced the BER and
power consumption to 74% and 85% as shown in Figure 3.18 and Figure 3.19
respectively. The proposed scheme has reduced an average of 0.005 BER/packet and
8 nano-joules power in comparison to base network model without ACH.
Minimum power consumption and high data rate in any wireless network is of
great importance for a given QoS. The wireless devices mostly have limited battery
power and the selection of appropriate air interface in ACH has proved the significant
advantage of the proposed technique. Reduced packet drop and BER by application
controlled handover technique not only enhance the bandwidth efficiency but also
reduce the number of re-transmission of packets which is directly proportional to the
power and complex error recovery techniques.
86
Figure 3.18: Network Average BER.
Figure 3.19: Network Average Power Consumption.
87
3.8 Conclusions
In this chapter it has shown how the development of novel Application
Controlled Handover (ACH) technique for FMC has been achieved. The modification
works in complementary with the standard WWIN architecture. Since most of the
modern wireless devices are mobile and battery-operated, it is important to detect the
device expected running time. This parameter could be used to enhance the network
functionality continuation by avoiding devices with short lifetime. Furthermore,
battery charger algorithm has been presented and simulated in order to approach the
real-life power model for the wireless battery-operated devices in this work.
In this chapter, transport layer enhancement (ACH) for the selection of
multiple radios according to application type, has been successfully proposed,
designed and implemented in terms of end to end network simulation. This chapter
concludes that an optical network with an Application Controlled technique would act
as a better convergence platform for future broadband network communication
systems. Handover algorithm not only enhanced the network overall performance but
also the reduction of packet drop up to 83%, BER to 74%, response time to 100% and
power consumption to 85%, which is an enormous achievement for any wired
wireless converged platform. On the other hand, there is decay in medium access
delay using proposed ACH.
The mesh and power controlled framework will be built on the ACH based
architecture, which is proposed by this work. A meshed and power controlled ACH
enhancement is proposed, discussed and compared in the next chapter.
3.9 References
[1] Evans, B.; Werner, M.; Lutz, E.; Bousquet, M.; Corazza, G.E.; Maral, G.; Rumeau,
R, “Integration of satellite and terrestrial systems in future multimedia
communications”, Wireless Communications, IEEE Vol. 12, No. 5, pp. 72–80,
Oct. 2005.
88
[2] Ali N. Zadeh, Bijan Jabbari, Raymond Pickholtz, Branimir Vojcic, "Self-
Organizing Packet Radio Ad Hoc Networks with Overlay (SOPRANO) ", IEEE
Communications Magazine, Vol. 40, No 6, pp. 149-157, Jun. 2002.
[3] Hongyi Wu, Chunming Qiao, Swades De, Ozan Tonguz, "Integrated Cellular and
Ad Hoc Relaying Systems: iCAR", IEEE Journal on Selected Areas in
Communications, Vol. 19, No 10, pp. 2105-2115, Oct. 2001.
[4] Ananthapadmanabha R., B. S. Manoj, C. Siva Ram Murthy, "Multi-hop Cellular
Networks: The Architecture and Routing Protocols", 12th IEEE International
Symposium, Vol. 2, pp. 78-82, Sep/Oct. 2001.
[5] Haiyun Luo, Ramachandran Ramjeey, Prasun Sinhaz, Li (Erran) Liy, Songwu Lu,
"UCAN: A Unified Cellular and AdHoc Network Architecture", proceedings of
9th ACM MobiCom, pp. 353-367, Sep. 2003.
[6] Xiaoxin Wul, S. H. Gary Chan, Biswanath Mukherjee, "MADF: A Novel
Approach to Add an Ad-Hoc Overlay on a Fixed Cellular Infrastructure ",
proceedings of IEEE WCNC, Vol. 2, pp. 549-554, Sep. 2000.
[7] Hung-Yun Hsieh, Raghupathy Sivakumar, "Performance Comparison of Cellular
and Multi-hop Wireless Networks: A Quantitative Study ", ACM SIGMETRICS,
MA, USA, pp. 113-122, June 2001.
[8] K. Jayanth Kumar, B.S. Manoj, C. Siva Ram Murthy, "MuPAC: Multi-Power
Architecture for Cellular Networks", proceedings of IEEE PIMRC, Vol. 4, pp.
1670-1674, Sep. 2002.
[9] John Bicket, Daniel Aguayo, Sanjit Biswas, Robert Morris, "Architecture and
Evaluation of an Unplanned 802.11b Mesh Network", proceedings of ACM
MobiCom, pp. 31-42, Sep. 2005.
[10] Roxana Zoican, Dan Galatchi, "Mobility in Hybrid Networks Architectures",
proceedings of IEEE TELSIKS, Vol. 1, pp. 273-276, Sep. 2005.
[11] Tonghong Li, Chan Kwang Mien, Joshua Liew Seng Arn, Winston Seah,
"Mobile Internet Access in BAS", proceedings of the 24th IEEE ICDCSW, pp.
736-741, Mar. 2004.
[12] R. Karrer, A. Sabharwal, E. Knightly, "Enabling Largescale Wireless Broadband:
The Case for TAPs", proceedings of HotNets, 2003.
89
[13] H.M.Chaskar, T.V.Lakshman and U Madhow , “ TCP over Wireless with link
level control: Analysis and Design Methedology”, IEEE Trans on Networking,
Vol. 7, No 5, pp. 605-615, Oct. 1999.
[14] IEEE Std 802.1Q-2005, IEEE standard for local and metropolitan area networks
virtual bridged local area networks, 2006.
[15] Jiancong Chen, Mixed Mode Wireless Networks: Framework and Power Control
Issues, Ph.D Thesis, Computer Science Department, Hong Kong University of
Science and Technology, May 2004.
[16] Wylie-Green,M.P, Ranta P.A, and Salokannel.J“Multi-band OFDM UWB
solution for IEEE 802.15.3a WPANs”, Advance in Wired and Wireless
Communication, IEEE/Sarnoff Symposium, pp. 102-105, April 2005.
[17] www.opnet.com (December 2010)
[18] Evans, M.; Hastings, N.; and Peacock, B. “Statistical Distributions”, 3rd edition
New York: Wiley, pp. 6-8, 2000.
[19] Rakhmatov, D.; Vrudhula, S.; Wallach, D.A.; “Battery lifetime prediction for
energy-aware computing”, Low Power Electronics and Design, 2002. ISLPED
'02. Proceedings of the 2002 International Symposium, pp. 154–159. 2002.
[18] Rakhmatov, D.; Vrudhula, S.; Wallach, D.A.; “A model for battery lifetime
analysis for organizing applications on a pocket computer “, Very Large Scale
Integration (VLSI) Systems, IEEE Transactions, Vol. 11, No. 6, pp. 1019 – 1030,
Dec. 2003.
[19] Benini, L.; Castelli, G.; Macii, A.; Macii, E.; Poncino, M.; Scarsi, R.; “Discrete-
time battery models for system-level low-power design”, Very Large Scale
Integration (VLSI) Systems, IEEE Transactions on, Vol. 9, No. 5, pp. 630 – 640,
Oct. 2001.
[20] Benini, L.; Bruni, D.; Mach, A.; Macii, E.; Poncino, M.; “Discharge current
steering for battery lifetime optimization “, Computers, IEEE Transactions on,
Vol. 52, No. 8, pp. 985 – 995, August 2003.
[21] Rao, R.; Vrudhula, S.; Rakhmatov, D.N.; “Battery modeling for energy aware
system design”, Computer, Volume 36, Issue 12, pp. 77 – 87, December 2003.
[22] Evans, M.; Hastings, N.; and Peacock, B. “Statistical Distributions”, 3rd edition
New York: Wiley, pp. 6-8, 2000.
90
[23] S R Chaudhry and H S Al-Raweshidy, “An application controlled mechanism for
heterogeneous Wired Wireless Integrated Network”, Wireless Personal
Multimedia Communications IEEE (IST 07), Budapest, Hungary, July 2007.
[26] HyangDuck Cho; JaeKyun Park; Keumsang Lim; Jongha Kim; Wooshic Kim,
“The mobile controlled handover method for fixed mobile convergence between
WLAN, CDMA and LAN”, Advanced Communication Technology, 2005,
ICACT 2005. The 7th International Conference, Feb. 2005, Vol. 1, Page(s): 559–
563.
[27] Nazem Khashjori, H.S. Al-Raweshidy, "Macrodiversity Evaluation in WCDMA
with Radio over Fibre access network", The Fifth IEEE International Conference
on Mobile and Wireless Communications Networks (MWCN2003), Singapore,
2003.
[28] Hamed Al-Raweshidy and Shozo Komaki , “Radio Over Fiber Technologies for
Mobile Communication Networks (Universal Personal Communications
Library)”, ISBN-10: 1580531482, Publisher: Artech House, 2002.
91
Chapter 4
Power Controlled Mesh Multi Radios
over Optical Fibre using ACH
The wireless-optical meshed network architecture has been proposed, as a
flexible solution to meet the ever-demanding needs of next generation hybrid
networks. In this chapter a meshed multiple radios approach that consists of 3G
(UMTS), WLAN, UWB, 802.15.4 and WiMAX over fibre has been implemented in
comparison to Application Controlled Handover (ACH) based network and simple
multi radio end-to-end wireless transmission. The proposed architecture implements
energy enhanced traditional distance based multihop mesh technique and maintains
the multi radio channels to provide real time quality of service continuity for
multimedia traffic.
4.1 Introduction Radio over Fibre (RoF) is a rather ideal technology for the integration of
wireless and wired networks. The main reason being is that it combines the best
attributes of two common communication methodologies. A wireless network
connection frees the end-user from the constraints of a physical link to a network,
which is a drawback of conventional fibre optic networks. Meanwhile optical
networks have an almost limitless amount of bandwidth with which to satiate even the
most bandwidth hungry customers where bandwidth for wireless networks can be a
significant bottleneck. Thus RoF networks offer customers the best of worlds by
allowing them to maintain their mobility while also providing them with the
bandwidth necessary for both current and future communication/entertainment
applications (i.e. HDTV, Video on Demand, 3DTV, video teleconferencing, etc.).
Such network topologies could be useful in places such as large buildings, subways
and tunnels where large amounts of people are mobile thereby making physical
connections impossible and standalone wireless systems being aced with the difficulty
92
of bandwidth limitations and handover issues.
While the concept of RoF networks has been around for several decades now,
there is still a substantial amount of work that needs to be undertaken in order to
improve their characteristics and capabilities. Several research priorities have been
identified that should be addressed in order to improve not only the capabilities of
RoF networks but also their cost effectiveness and robustness [1]. The requirement to
communicate a very high data rate such as rich multimedia with low power and low
cost are the reason behind the integration between wireless and wired domains using
RoF as this integration is the only way to allow high data rate with low power and low
cost. The integration has the potential to make the networks more transparent,
dynamic, faster and greener. Reconfigurable enhanced mesh architecture for multi-
application, multi-homed and multi-service supported mobile device is presented to
enable seamless mobility across different access technologies including wired and
wireless infrastructures.
In an integrated 3G+WLAN+WPAN+WiMAX architecture as illustrated in
Figure 4.1, more data and multimedia applications are carried end-to-end over the
current Internet protocol infrastructure. Fundamentally Wired Wireless Integrated
Networks (WWIN) will reshape the telecommunications industry. Technological
changes in the telecommunication industry have not only led to the creation of new
trends in the deployment of next generation networks, but also the convergence with
information technology sector. In particular, the so-called WWIN has become a
reality due to market demand and industrial competition. Fixed Mobile Convergence
(FMC) enables users to work with existing wired and wireless networks. 3G has been
under global deployment and it accesses the network by using the cellular and
WiMAX networks. WLAN and UWB are free of charge high data rate
communication air-interfaces. WLAN, UWB, WiMAX and UMTS can serve the
Internet core network via the Transmission Control Protocol/Internet Protocol
(TCP/IP). TCP/IP enables users to adopt client server communication by accessing
the individual network.
93
Figure 4.1: Proposed Power Controlled Mesh Multi Radio Network. A. Research Challenge
The work in this chapter is the enhancement and realisation of the proposed
architecture for next generation multi radio wired and wireless converged networks
and enhanced power efficient scheme for multi radio mesh networks, which provide
data transmission and service continuity in the WWIN environment. The proposed
mesh architecture uses a cross layer design to operate on the transport layer between
the wired and wireless network infrastructures. The multi radio mesh devices have
installed power efficient scheme and move freely in the wireless domain. The
handover provides service continuity and data transmission by selecting the wireless
technology depending on the required power. The traffic was recorded at the remote
94
server that was located in the wired network and it was transported from devices that
were located in the wireless network as presented in Figure 4.1.
This chapter is organised in 7 sections. Section 4.2 outlines the related work in
the era and hierarchy of the classification. Section 4.3 presents the brief architecture
of proposed multi radio mesh over fibre architecture and organisation of
methodology. In Section 4.4, the development and design of core network has been
discussed in details. Section 4.5 demonstrates the network scenarios, implication of
scenarios and layout for the proposed architecture. Section 4.6 elaborates the results
and discussions leading to precise conclusions in Section 4.7.
4.2 Global Contribution and Related Work
In future Internet where wired and wireless networks are integrated, the traffic
volume of wireless users should expand greatly and the QoS for wireless users will be
one of the most important technical issues. In the wireless network environment, TCP
is one of most important protocol, suffers from significant throughput degradation due
to the lossy characteristics of a wireless link. Therefore, in order to design the next
generation networks, it is necessary to know how much improvement a WWIN
brings. The throughput of a TCP connection is dependent upon both packet loss and
application response delay. Several papers have been published in the domain of
wireless TCP, for example, reference [2] evaluates throughput performance of a
single wireless TCP connection experimentally, and reference [3] analyses the
throughput performance mathematically using a fluid flow model. These papers only
dealt with wireless TCP and did not take care of the interaction of wireless and wired
TCP in the FMC environment.
The architecture classification for WWIN is based on: Intracellular, Inter-
cellular and Extra-cellular architectures as shown in Figure 4.2. In Intracellular
architectures, each Access Point or Base Station (AP/BS) communicates with any
mobile node (MN) in its coverage area in a multi-radio mode. On the other hand, in
Inter-cellular architectures, each AP/BS communicates with any MN in the coverage
area of other APs/BSs in a multi-radio mode. Finally, Extra-cellular architectures
enable an AP/BS to communicate with MNs that are not in the coverage area of any
95
AP/BS. Any of these classifications may employ dedicated radio stations.
Furthermore, Intra/Inter cellular ad hoc mode architectures can be classified as multi-
radio systems in which MNs act either in single-radio mode or multi-radio mode.
Figure 4.2 illustrates this classification with respect to the well-known references such
as [4]-[18] in this research era.
Figure 4.2: Classification of Wired Wireless Converged Network Architectures.
4.3 Proposed Network Framework
The existing integrated high-speed wired and wireless mesh networks are
shown in Figure 4.1 that reveals the general constitution of proposed architecture.
Both wired and wireless networks have separate service platforms. FMC network
supports data service in the existing voice/data based wireless infra, and the existing
high bandwidth wired network infrastructure.
96
The proposed FMC simulation architecture is based on energy efficient multi-
radio mesh such as UMTS, WiMAX, WLAN, UWB, ZigBee and optical fibre
networks. In this chapter a novel FMC environment is proposed by using converged
TCP/IP and a transport layer between wired and wireless networks. The proposed
architecture uses the transport layer and TCP/IP for each network constituting FMC.
In this mesh architecture for energy efficiency, mobile devices select next hop
according to the distance instead of random nodes. The hybrid interface selection
depends on the access of HTTP data using the UMTS cellular networks. For FTP and
real time database applications, such as remote account management and payroll
systems for a multinational corporate, WLAN and WiMAX is a widely used solution.
For real time video calling and video conferencing UWB is prominent candidate
supporting 200 Mbps [2]. [3]. WiMAX has a shared bandwidth up to 72Mbps [3].
Figure 4.1 shows various mobile networks in the wireless network domain
having access to UMTS, WLAN, WiMAX, ZigBee and UWB air interfaces. The
network protocol stack can communicate in a cross layer fashion from PHY to APP
Layer to detect the transmission of data, handover to the suitable air-interface,
distance measurements and send it to the remote server. Once the server gets the
request, it records real time data transmitted from user’s device.
4.4 Development of Core Network Components
This section presents an energy efficient multi radio over fibre components for
wired and wireless integrated networks. The heterogeneous mobile devices not only
sense the signal strength but also determine the closest relay node if the Radio Access
Unit (RAU) is at distance. A soft handover has been considered in the simulation
network due to the session handling of connections of more than two radios e.g.
connections to UMTS, WLAN, WiMAX and UWB can be maintained by one device
at the same time. It is a trade off between the one channel occupancy (hard handover)
and the QoS in terms of soft handover’s make-before-break (session continuity). This
section elaborates the wireless and optical components for proposed network protocol
stack, the design of network components (wired and wireless) and finally the
mathematical process model for an energy efficient mesh.
97
The proposed components perform the following functionality:
1. Cross Layer Network Protocol Stack
2. Multi Radio Network Connection Schematic
3. Multi Radios over Fibre Uplink/Downlink Design
4. Meshed Network Architecture
5. Distance Controlled Power Efficient Mesh
A. Cross Layer Network Protocol Stack
Figure 4.3 depicts the protocol stack for the proposed architecture. Seamless
bridging is supported natively by this protocol, which enables both information and
process data to be easily exchanged in heterogeneous environment. The multi radio
transportation manager adopts the lower layers of the protocol stack (PHY). Despite
the relatively low speed of UMTS (384 kbps), it is able to guarantee deterministic
behaviour. As long as the load on the network is well below the theoretical available
bandwidth, the proposed architecture guarantees that any low or high data is
transmitted to the server within the suitable application response time.
Figure 4.3: Proposed Multi Radio Network Protocol Stack. B. Multi Radio Network Connection Schematic
The design of components for mobile devices is illustrated in Figure 4.4.
(I) The Data Control component that can exchange data with external device,
(II) The Application and Air Interface signal acquisition component that can
98
receive the radio signal strength of the wireless network,
(III) The Socket Control component that controls the communication resources
inside the device with the wireless network platforms according to the demand
of quality data transmission requirements.
Figure 4.4: Multi Radio Network Connection Control.
The proposed wireless mesh mode used virtual sockets for all multiple air
interfaces such as UMTS, WLAN, WiMAX, ZigBee and UWB [19]. [20]. In this
chapter a design of five virtual sockets for the five radios have been proposed and
developed. In the Figure 4.4, the signal control component can control all the five
radio interfaces between data control and the socket control component. The
application and air interface acquisition component provides the signal strength of
AP/BS. The network controller is connected to the multi radio transportation
manager.
C. Multi Radios over Fibre Uplink/Downlink Design
The general optical network components are depicted in Figure 4.5. The
uplink where packets are transmitted from and downlink signals are separated using a
RF circulator as shown in Figure 4.5b and Figure 4.5c. A single mode fibre (SMF) is
99
used with different wavelength for uplink and downlink. To eliminate the effect of the
wireless channel on the system performance an RF cable connects the RAU. The
uplink and downlink RF signals are separated using a circulator. The proposed design
of uplink and downlink are shown. The electrical signal is converted to an optical
signal using a variable bandwidth distributed feedback (DFB) laser. An optical
combiner is used which allows a single length of SMF to be used for both directions.
To ensure that the optical power levels in the system were constant, a bidirectional
optical attenuator was employed for each fibre length. At the receiver end, an RF
amplifier is used to increase the radio signal power level, which has a maximum value
of 100mW in Europe [20]. It should be noted that all amplification occurs at the RAU.
Figure 4.5 a: Multi Radios Design.
Figure 4.5 b: Multi Radios Downlink.
100
Figure 4.5 c: Multi Radios Uplink. D. Meshed Network Architecture
In a meshed network, there are mesh devices (MDEVs) and mesh network
coordinators (MNCs) in addition to the legacy 802.15.3 devices and Communication
between MDEVs that belong to different independent networks is possible in a mesh
configuration. The MNCs can communicate with each other and they can also manage
their own networks that can contain the MDEVs . Thus both inter-network and intra-
network communication can take place in meshed structure as shown in Figure 4.6.
Figure 4.6: Mesh Network Architecture.
101
Unlike the intra-network communication that takes place in the legacy single hop
network in which a single network coordinator coordinates the channel time in the
superframe, several MNCs share the channel time in the superframe among them. For
this reason, the superframe is divided into Medium Access Slots (MASs) that are of
equal length. The length of the superframe is kept the same throughout the mesh
network and all the MNCs synchronize their beacons to a common reference MNC.
At the front end, proposed architecture consists of a multi-hop wireless mesh
network while at the back end an optical access network provides connection to the
Internet. At the back end the dominant technology is the passive optical network
(PON) having optical line terminals (OLTs), located at the central office (CO), and
optical network units (ONUs) to provide connection to wireless gateway routers.
Different PON segments are supported, with each segment radiating from a single
OLT at the CO to multiple ONUs near end-users. The PON interior elements are
basically passive combiners, couplers and splitters. Since no active elements exists
between the OLTs and the ONUs the PONs are considered robust networks. In
traditional time-division-multiplexed (TDM) PONs an upstream and a downstream
wavelength channel is used for bidirectional communication as shown in Figure 4.5.
The network architecture is illustrated in Figure 4.1. To provide connection between
PONs a reconfigurable optical backhaul, allowing easy bandwidth reallocation, can be
used.
Although having low installation and maintenance cost, due to the passive
infrastructure, in the traditional PON, all end users share the CO. That is, a packet
sent to a specific user, connected to a particular ONU, is broadcast to all ONUs. As
users demand for more bandwidth, wavelength division-multiplexing (WDM) PONs
may become a better alternative. A WDM PON solution supports multiple
wavelengths such as WiFi, ZigBee, UWB, WiMAX and UMTS over the same fiber
infrastructure, but active components become necessary. Incorporating WDM in a
PON allows the support of much higher bandwidth and scalability when compared to
the standard PON, which operates in single-wavelength mode.
Regarding the wireless infrastructure, standard WiFi or WiMAX technology
can be used for wireless mesh connectivity. These wireless access technologies have
102
been employed worldwide. Several wireless routers provide multihop connectivity for
user traffic delivery toward a few wireless gateway routers that are connected to the
ONUs of the optical back end. An ONU can drive multiple gateways. An individual
user scattered over such geographical area will associate with a nearby wireless router
for Internet access. A mesh router can gain access to a gateway router through
multiple paths. In the upstream direction traffic can be delivered to any of the
gateways while in the downstream direction traffic is sent to a specific wireless router.
In such multi-radio networks each radio interface has the capability of
switching over orthogonal channels and transmission or reception is possible at any
channel at a time [21]. That is, considering a specific channel, only links that are
located out of their mutual interference range can transmit at the same time.
Consequently, a careful channel assignment becomes critical in such networks so that
more simultaneous transmission occurs, leading to a global system capacity
improvement [22].
E. Distance Controlled Power Efficiency
Multiple power efficient management schemes in wireless networks have been
proposed in [23]. Lower transmission power leads to lower interference level and
better end-to-end throughput. Power saving in multi radios has been studied in order
to increase the mobile nodes operating time [24]. In a multi radio multi hops network,
wireless network protocols can manage power consumption for each individual node
and the total power distribution in the entire network. Figure 4.7a shows the source
data is node (A), while destination is node (B) in a mesh domain.
In this case, the destination node (B) is still in communication range of the
data source node (A), but it requires high transmit power to meet the necessary signal
quality. Instead, the node (A) can use the multi-hop feature of ad hoc networks and
build a route from (A) to (B) via an intermediate node (C), where (C) is located in
midway between (A) and (B) as shown in Figure 4.7a. The decision to route data from
(A) to (B) via (C) should take into consideration the battery level in both nodes (A)
and (C) as well as link stability.
103
Figure 4.7a: Power Controlled Transmission.
Figure 4.7b: Distance based Power Controlled Schema.
I. Power Saving using Multiple Hops
Power level is a scale for the minimum power required by the sender node to
reach the next hop. If the next hop can be reached via another node by lower
transmission power, then this new path should be adapted. Figure 4.7a illustrates this
concept.
For node (A), less transmission power is required to reach the destination node
(B) via intermediate node (C) instead of sending data directly from (A) to (B). The
decision to change the route from direct link between (A) and the destination node (B)
to an indirect route via (C) is taken by calculating the battery life of node (C) and the
total throughput and efficiency of the new route.
104
In the source node:
1. When (battery / expected use time) = threshold: Then start transmission save
mode
2. Check all the next hop neighbour nodes about a possible volunteer (the old
destination node should be included in this message)
3. If a neighbour node can reach the old destination and source-neighbour required
transmission power is less than the original; then a Volunteer Request message
will be sent back to the source node
4. The route will be changed from the old destination to the new volunteered
neighbour.
In the neighbour node:
1. When a Volunteer request message is received:
2. If the battery / expected use time > threshold: Then start Volunteer Process
3. If the old destination is in reasonable reach of transmission: Then send Volunteer
Acknowledge message.
II. Transmission and receiving power
In Figure 4.7b, node “A” can reach node “C” directly with power . However,
node “C” could be reached indirectly via node “B”. The second path power is the sum
of two parts ( and ). The wireless nodes transmit data to a certain range
around the data source. The coverage range is a function of the transmission power
and the medium coefficient. Equation (4.1) calculates the received power.
(4.1)
where:
105
: Received power.
: Log-normal coefficient
: Distance between the source and destination
: Transmitted power.
Received power values (of transmitted signals from source nodes) are
assumed to be equal at the destination nodes. Furthermore, the medium coefficient
is the same for all data transmissions since it is the same ambience as shown in
Equation (4.2).
(4.2)
where:
K is a constant value.
(4.3)
where:
: distance between “A” and “C”
: distance between “A” and “B”
: distance between “B” and “C”
Based on these two assumptions, the transmission power of the two paths could be
calculated as the following:
The direct path power:
(4.4)
Based on Equation (4.4), could be replaced as:
PACdirect = K dAB + dBC( )4 (4.5)
106
(4.6)
The indirect path power:
and (4.7) From Equation (4.7), the required power for the indirect path between “A” and “C” is:
(4.8)
(4.9)
Comparing Equation (4.6) with Equation (4.9), the direct transmission power is higher
by:
(4.10)
From Equation (4.10) it is evident that the multiple-hops path requires
less power than the direct path. The proposed mathematical model has been
implemented to simulate the impact of power efficiency in a multi radio mesh
environment.
4.5 Network Setup and Layout
The simulation model in OPNET [25] has been divided into three major
scenarios (I) multiple radios in end-to-end wireless domain (wireless base model), (II)
multiple radio over fibre network (III) power efficient mesh enabled multiple radio over
fibre network (meshed multi radios over fibre). All three scenarios are simulated in
parallel and compared on the basis of same parameters; the details are in the
following sub sections.
The Figure 4.1 represents the multi radio mobile devices capable of WLAN,
UWB, WiMAX, ZigBee and UMTS connectivity. The Radio Access Units (RAUs)
107
supports all MAC and PHY layers of WLAN, WiMAX, ZigBee UWB and UMTS. The
UWB model used here is 10 meters in range with a data rate up to 200 Mbps. This
wide network was setup and simulated to analyse the scope of the simulation and to
achieve reasonable simulation time of 60 minutes.
A. Wireless Base Model In this scenario, the behaviour of five wireless technologies UMTS, WLAN,
WiMAX, Zigbee and UWB were examined within the framework of a pure wireless
domain. Firstly, the mobile device communicates to the remote server using a simple
http application with UMTS infrastructure. The data traffic changed from HTTP to
FTP and then to Video conferencing respectively. The WLAN, WiMAX and UWB
were accessed via RAUs, which were connected to a wireless backbone server. An
IP gateway (i.e. an enterprise router) connected the wireless network to an IP cloud
used here to represent the backbone Internet connectivity. The remote server
supports three kinds of traffic FMC such as HTTP, FTP and real time video
calling/conferencing.
B. Multiple Radios over Fibre Model using ACH In the second simulation scenario, setting and configurations are same as for
the first one. The only addition is of the implementation of multi RoF. Mobile devices
communicate with radio access points with no hoping. The RAUs are connected with
Internet cloud using optical fibre (10 Gbps) simulating a real life environment [26]-[28].
C. Power Controlled Meshed Multiple Radios over Fibre Model The third scenario keeps entire simulation, the setting and configurations
same as for the second scenario. The functionality addition is that the mesh enabled
multi RoF concept is introduced with proposed energy efficiency. Mobile devices
communicate with radio access points with distance based hoping to minimize the
power transmission.
108
Figure 4.8 is implemented and carried out. The technique in each multi radio
node calculates the set of parameters for all possible destinations at periodic time
intervals. In each node, the adjacent neighbour nodes are listed in a table stored with
the minimum required power to reach next closest node. Source routing is used at
the nodes according to a QoS set of requirements, which specifies the optimization
function to be used for the selection of the paths. When a data packet is generated at
the source node, the node applies the function to the cost vectors (check destination
node, check battery and power threshold for transmission) of the paths to select the
optimal path. If the source node battery power level is over threshold, the packet is
transmitted to the destination in peer-to-peer mode. If the battery power is under
threshold value, the algorithm looks for a lower cost node with minimum required
power in neighbourhood and forwards it the data packet.
Figure 4.8: Power Controlled Mesh Work-Flow.
109
The proposed architecture made the data exchange very bandwidth efficient
and enhanced system performance as discussed and presented in Section 4.6. The
simulation network traffic type and parameters for three scenarios are described in
Table 4.1 and Table 4.2 respectively.
Table 4.1: Network Application Traffic Profiles APPLICATIONS
H= Heavy L=Light
Multi Radio Network Scenarios
HTTP
FTP
Real Time-Video Conf.
Wireless Base Model L
H
H
Multi Radio over Fibre L
H
H
Meshed Multi Radio over Fibre
L
H
H
Table 4.2: Network Simulation Environment
4.6 Performance Evaluation and Discussions
This section carried out an imperative performance evaluation for mesh
enabled multiple radios over fibre network with a special focus on the performance of
the UMTS, WiMAX, WLAN, Zigbee and UWB when subjected to rich multimedia
applications. Real time traffic was introduced in the network in the form of video
conferencing. Video conferencing encompasses data, voice and images and
adequately represents the ultimate heavy multimedia traffic. No extensive network
management techniques were configured but the proposed architecture has been used
to identify the effect of multiple radio interfaces on the performance of the optical
fibre network.
Number of Nodes
25
Simulation Area
500 x 500 Meters
Simulation Time
60 Min
Fibre Channel bit rate
10 Gbps
110
The performance measurements that are relevant to network traffic, such as
wireless and cellular link utilisation in terms of throughput were obtained and
discussed. More specifically, the response times of applications (HTTP, Database and
Video Conferencing) at the mobile devices were also compared to ascertain the user
with expected QoS. Additionally, some critical network parameters such as media
access delay, end-to-end packet drop, bit error rate and network power consumption
have also been evaluated for both scenarios. A simulation time of one hour was set for
both scenarios to achieve feasible results.
4.6.1. Network Throughput
The results showed that a higher network throughput has been achieved for the
meshed multi radios over fibre. Figure 4.9 presents a comparison of average network
throughput for same type data traffic of three scenarios. It shows meshed multi radio
over fibre has 50% increase in the throughput of the networks because of multi-
hoping and relay devices. The application controlled optical network has reached up
to an average of 12 Mbps as compared to 5 Mbps by the network without mesh mode.
Figure 4.9: Network Average Throughput.
111
4.6.2. End-to-End Network Delay
Figure 4.10 presents the end-to-end network delay that indicates the
comparison among wireless base model, multi RoF and meshed multi RoF models.
The network average delay has been increased to about 50% when the proposed
architecture is applied.
The base model performed best in comparison to multi RoF and meshed multi
RoF model as compared to multihop communication, which lead to increased delay.
This result is expected when the proposed approach is applied due to RAU’s and
multi hoping. The end-to-end delay for multi RoF and meshed multi RoF is 15 ms as
compared to 1 ms delay in base model.
Figure 4.10: Network Average End-to-End Delay. 4.6.3. Network Multiple Media Access Delay
In previous results, it has been observed that wireless base model perform
better in terms of end-to-end delay. The proposed architecture works on the transport,
MAC and PHY layers. It reduces the medium access delay of the network because of
availability of multiple air interfaces, whereas with base network it resulted in 0.8
seconds more delay as shown in Figure 4.11.
112
Figure 4.11: Network Average Medium Access Delay. 4.6.4. Packet Drop
Figure 4.12 shows the peak-to-peak network average packet drop comparison
for the network scenarios. The base network shows very non- linear behaviour
because of high variation in transmission. It has dropped up to 10 packets/second and
gradually reduced to an average of 5 packets/second. On the other hand, meshed and
multi RoF models behaved well by reducing the packet drop up to 53%.
Figure 4.12: Network Average Packet Drop.
113
4.6.5. HTTP Response Time
There was great improvement in the HTTP response time when proposed
architecture was applied as shown in Figure 4.13. The result shows that the meshed
and multi RoF models can handle higher loads than the base model due to the higher
bandwidth of fibre. It can be seen clearly from Figure 4.13 that both RoF network
models performed well whereas, the response time reached 4.5 seconds for the base
model.
The meshed and multi RoF performed excellently and reached an average response
time of 0.4 to 1.3 seconds respectively.
Figure 4.13: HTTP Average Response Time. 4.6.6. FTP Response Time
An improved performance is also seen in FTP during the simulation. A heavy
FTP needs higher authentication and security information compared to a simple
HTTP application. Therefore it enhanced the load on the network. The FTP response
time reached upto 10 seconds for the base model as shown in Figure 4.14.
However, proposed RoF architecture reduced response time to an average of 0.2
and 0.3 seconds for multi RoF and meshed multi RoF respectively, which is 100%
less then base model.
114
Figure 4.14: FTP Average Response Time.
4.6.7. Video Conferencing Response Time
The base model for real time video conferencing shows unacceptable response
time of 27 seconds because heavy multimedia video transmission overloaded the
radio link. The proposed RoF enhanced the response time from the video server. An
average improvement of 23 packet/sec for video transmission has been seen in Figure
4.15, when proposed concept applied.
Figure 4.15: Video Conferencing Average Response Time.
115
4.6.8. End-to-End Power Consumption
The base network achieved higher power consumption. The glitches in the
graphs are because of the variation in data traffic and the high bit rate on real time
application demand. It can be seen that multi RoF has reduced the power consumption
to 7% as shown in Figure 4.16. The proposed distance based power efficient mesh
mode reduces the power consumption to 8% further as compared to multi RoF model.
It has reduced an average of 10 mW power in comparison to the base model.
Figure 4.16: Network Average Power Consumption.
4.7 Conclusions
Power controlled meshed multi radio over fibre using ACH has been presented
with a novel architecture for wired wireless synergy network. In this chapter,
implementation of multiple radios with and without mesh mapping, has been
successfully proposed, designed, implemented, discussed and simulated.
This chapter concludes that a meshed multi technique would act as a better
convergence platform for future broadband network communication systems. In
comparison to the traditional network base model, the proposed layout not only
enhanced the network overall performance but also reduced packet drop upto 53%,
application response time upto 100% and power consumption by 15%, which is very
116
efficient for any wired wireless converged platform.
4.8 References
[1] Evans, B.; Werner, M.; Lutz, E.; Bousquet, M.; Corazza, G.E.; Maral, G.;
Rumeau, R, “Integration of satellite and terrestrial systems in future multimedia
communications”, IEEE Wireless Communications, vol. 12, no. 5, pp. 72 – 80, Oct.
2005.
[2] Wylie-Green,M.P, Ranta P.A, and Salokannel.J“Multi-band OFDM UWB
solution for IEEE 802.15.3a WPANs”, Advance in Wired and Wireless
Communication, IEEE/Sarnoff Symposium, pp. 102-105, April 2005.
[3] S R Chaudhry, H S Al-Raweshidy and S S A Obayya, “A comparative study of
MB-OFDM and DS-CDMA in UWB WPAN”, WWRF Nov 2004.
[4] Ali N. Zadeh, Bijan Jabbari, Raymond Pickholtz, Branimir Vojcic, "Self-
Organizing Packet Radio Ad Hoc Networks with Overlay (SOPRANO) ", IEEE
Communications Magazine, vol. 40, no 6, pp. 149-157, Jun. 2002.
[5] Hongyi Wu, Chunming Qiao, Swades De, Ozan Tonguz, "Integrated Cellular and
Ad Hoc Relaying Systems: iCAR", IEEE Journal on Selected Areas in
Communications, vol. 19, no. 10, pp. 2105-2115, Oct. 2001.
[6] Ananthapadmanabha R., B. S. Manoj, C. Siva Ram Murthy, "Multi-hop Cellular
Networks: The Architecture and Routing Protocols", 12th IEEE International
Symposium, vol. 2, pp. 78-82, Sep/Oct. 2001.
[7] Haiyun Luo, Ramachandran Ramjeey, Prasun Sinhaz, Li (Erran) Liy, Songwu Lu,
"UCAN: A Unified Cellular and AdHoc Network Architecture", proceedings of 9th
ACM MobiCom, pp. 353-367, Sep. 2003
[8] Xiaoxin Wul, S. H. Gary Chan, Biswanath Mukherjee, "MADF: A Novel
Approach to Add an Ad-Hoc Overlay on a Fixed Cellular Infrastructure ",
proceedings of IEEE WCNC, vol. 2, pp. 549-554, Sep. 2000.
[9] Hung-Yun Hsieh, Raghupathy Sivakumar, "Performance Comparison of Cellular
and Multi-hop Wireless Networks: A Quantitative Study ", ACM SIGMETRICS, MA,
USA, pp. 113-122, June 2001.
[10] K. Jayanth Kumar, B.S. Manoj, C. Siva Ram Murthy, "MuPAC: Multi-Power
Architecture for Cellular Networks", proceedings of IEEE PIMRC, vol. 4, pp. 1670-
1674, Sep. 2002.
117
[11] John Bicket, Daniel Aguayo, Sanjit Biswas, Robert Morris, "Architecture and
Evaluation of an Unplanned 802.11b Mesh Network", proceedings of ACM
MobiCom, pp. 31-42, Sep. 2005.
[12] Roxana Zoican, Dan Galatchi, "Mobility in Hybrid Networks Architectures",
proceedings of IEEE TELSIKS, vol. 1, pp. 273-276, Sep. 2005.
[13] Tonghong Li, Chan Kwang Mien, Joshua Liew Seng Arn, Winston Seah,
"Mobile Internet Access in BAS", proceedings of the 24th IEEE ICDCSW, pp. 736-
741, Mar. 2004.
[14] R. Karrer, A. Sabharwal, E. Knightly, "Enabling Largescale Wireless Broadband:
The Case for TAPs", proceedings of HotNets, 2003.
[15] V. Angelakis, M. Genetzakis, N. Kossifidis, K. Mathioudakis, M. Ntelakis, S.
Papadakis, N. Petroulakis, V.A. Siris, "Heraklion MESH: An Experimental
Metropolitan Multi-Radio Mesh Network". In Proc. of 2nd ACM Int'l Workshop on
Wireless Network Testbeds, Experimental evaluation and CHaracterization
(WiNTECH'07), held in conjunction with ACM MOBICOM'07, Montreal, Canada,
September 2007.
[16] Isabelle Siaud, Anne-Marie Ulmer-Moll, N. Malhouroux-Gaffet, V. Guillet:
Chapter 18, "An Introduction to 60 GHz Communication Systems: Regulation and
Services, Channel Propagation and Advanced Baseband Algorithms", Wiley, Short-
Range Wireless Communications: Emerging Technologies and Applications, ISBN:
0470699957, March 2009.
[17] https://www.ict-smartnet.eu/ (Dec 2010)
[18] O. Sallent, L. Giupponi, J. Nasreddine, R. Agustí and J. Pérez-Romero,
“Spectrum and Radio Resource Management in Cognitive Heterogeneous Wireless
Networks”, IEEE Vehicular Technology Magazine, December 2008.
[19] S R Chaudhry and H S Al-Raweshidy, “An application controlled mechanism for
heterogeneous Wired Wireless Integrated Network”, Wireless Personal Multimedia
Communications IEEE (IST 07), Budapest, Hungary, July 2007.
[20] HyangDuck Cho; JaeKyun Park; Keumsang Lim; Jongha Kim; Wooshic Kim,
“The mobile controlled handover method for fixed mobile convergence between
WLAN, CDMA and LAN”, Advanced Communication Technology, 2005, ICACT
2005. The 7th International Conference, vol. 1, pp. 559–563, Feb. 2005.
118
[21] Jihui Zhang, et al., “Joint routing and scheduling in multi-radio multi-channel
multi-hop wireless networks”, BROADNETS, Vol. 1, pp. 631-640, 2008.
[22] Krishna N. Ramachandran, et al., “Interference-Aware Channel Assignment in
Multi-Radio Wireless Mesh Networks”, IEEE INFOCOM, pp. 1-12, 2006.
[23] E.-S. Jung and N. H. Vaidya. An Energy Efficient MAC Protocol for Wireless
LANs. In INFOCOM, June 2002.
[24] E.S. Jung and N.H. Vaidya, “A power control MAC protocol for ad hoc
networks,” Wireless Networks, vol.11, no.1, pp.55–66, 2005.
[25] http://www.opnet.com/ (Dec 2010)
[26] Nazem Khashjori, H.S. Al-Raweshidy, "Macrodiversity Evaluation in WCDMA
with Radio over Fibre access network", The Fifth IEEE International Conference on
Mobile and Wireless Communications Networks (MWCN2003), Singapore, 2003.
[27] Hamed Al-Raweshidy and Shozo Komaki, “Radio Over Fiber Technologies for
Mobile Communication Networks (Universal Personal Communications Library)”,
ISBN-10: 1580531482, Publisher: Artech House, 2002.
[28] S R Chaudhry and H S Al-Raweshidy, “Application-controlled handover for
heterogeneous multiple radios over fibre networks”, vol. 2, no10, pp. 1239-1250, IET
Communications, November 2008.
119
Chapter 5
Intelligent Application Priority
Assignment Mechanism for BWSN __________________________________________________________
Biomedical wireless sensor networks (BWSNs) need to
guarantee reliable and timely transfer of data in emergency situations. This chapter
presents a novel data transmission scheme based on Application Controlled Handover
(ACH) mechanism, named as *Intelligent Application Priority Assignment
Mechanism (IAPAM) for BWSNs. Current schemes use static/fixed priority
assignment mechanism based on source of data and not on the criticality and
importance of the data in different situations. IAPAM dynamically schedule different
types of data flows based on their time critical nature. IAPAM smartly assigns priority
to individual data packets rather to particular service or flow by continuously
monitoring queuing delay providing end-to-end QoS without invoking any congestion
control and avoidance mechanism. Experimental results show that IAPAM performs
better than current standard of BWSN in terms of throughput and end-to-end delay for
time crucial medical applications.
5.1 Introduction
During the last 5 years, various healthcare solutions have been proposed using
information and communication technologies (ICT) [1]. Patient monitoring systems
based on Biomedical Wireless Sensor Networks (BWSNs) and body area networks
(BANs) has improved patients' health status during chronic care [2]-[4], thus enabling
physicians to intervene more timely in emergency situations [5]. Patient monitoring in
chronic cases demands pervasive healthcare for efficient response to treatment [6],
[7].
* Intelligent: The word “Intelligent” is used in the context of enhanced and smart proposed mechanism.
120
A BWSN monitors important medical data such as body temperature (Temp),
blood pressure (BP), electrocardiogram (ECG), electroencephalograph (EEG),
electrooculography (EOG), electromyogram (EMG), heart rate, breathing rate and
transmits it to a medical center, where this data is used for health status monitoring,
medical analysis and emergency treatments [8]-[10]. A BWSN needs to guarantee that
the packets having time critical data can be sensed and delivered to a medical center
reliably and efficiently within the delay bounds. A routing protocol is an essential part
of BWSN to route the time critical data with minimum end-to-end delays using
congestion free intermediate nodes. BWSN requires further research on QoS issues
realizing requirements of time critical data flows in medical scenarios.
Time critical data flows in medical scenarios handle data
according to time constraints associated with the flow and the urgency of data. A
BWSN requires an efficient data priority assignment and scheduling mechanism to
accomplish this task. It also needs to continuously monitor network conditions to use
an optimal path for data delivery.
In this chapter, an Intelligent Priority Assignment Mechanism
(IAPAM) has been proposed by authors to increase throughput of time critical data
packets. IAPAM decides at runtime which data flow is to be given high priority based
on its urgency. The data generated by the source sensor nodes is compared with their
respective threshold values. If the value of any type of sensed data is greater or close
to its threshold value, priority is assigned to that data stream. At the intermediate
sensor nodes a classifier is used. The task of the classifier is to classify different data
flows and assign queues according to their priority and the source node identity. A
scheduler adaptively schedule packets on the basis of the priority assigned by the
IAPAM. At each intermediate node, an average queuing delay is calculated for both
priority and non-priority queues. The calculated value is compared against threshold
and appropriate action is taken in case the value is greater than the threshold. The
proposed scheme is implemented and simulated in JSIM network simulator [11], [12].
Previous schemes [8]-[12] assign priority to the data flows on
the basis of the source from which the data packet is generated. When there are
121
different data flows and priority is to be assigned to the packets on the basis of their
urgency, in that case a mechanism is needed which monitors QoS among different
data flows and within those data flows. In IAPAM priorities are assigned to the
packets on the basis of their data contents and not on the basis of source nodes as in
other schemes.
The remainder of this chapter is organized as follows: Section
5.2 consists of contribution and related work from academia and industry. Section 5.3
presents the challenges, problem and methodology to develop the IAPAM. Section
5.4 provides related detailed description of IAPAM scenario in BWSN. Section 5.5
describes IAPAM functional components, enhancements in terms of intelligence and
their working architecture. Section 5.6 presents performance analysis of IAPAM with
brief discussion of all crucial parameters. The chapter is concluded in section 5.7.
5.2 Global Contribution and Related Work
Patel et al. present results evaluating possible use of wireless sensor networks
for activity analysis of Parkinson's patients [13]. In this paper, motor fluctuations
indicate the effectiveness and timing of medications for Parkinson's patients. In 14,
authors use inertial sensors for monitoring of a distributed fall detection system.
Boyle et al. estimates the respirations rate using existing cardiac sensor
reducing the number of sensors to enhance user's convenience [15]. This simplifies
the cumbersome task of monitoring multiple physiological parameters in healthcare
systems. Yoo et al. present a cardiac monitoring system with planner-fashionable
circuit board shirt reducing number of sensor set [16].
I. Martı´nez et. al. presents an end-to-end based patient monitoring solution
using ISO/IEEE 11073 (X73) in the bedside environment and EN13606. This plug-
and-play sensor network communicates with a gateway that gathers medical
information and sends the data to a monitoring server. The monitoring server
transforms this information to an EN13606 extract to be stored on the electronic
healthcare record (EHR) server. The system complies with the last X73 and EN13606
122
available versions providing end-to-end solution [17].
X. Liang and I. Balasingham [8] discussed cross layer QoS aware routing
framework for biomedical sensor networks. Packets are classified as data packets and
control packets having different priorities. When packets in a buffer reach a threshold,
the proposed framework communicates with the user application to reduce the service
rate. This scheme is not dynamically adaptive to the network changes, and it requires
user intervention.
F. Xia and W. Zhao [18] presented a fuzzy logic based controller scheme to
provide QoS management in wireless sensor networks. The idea is to regulate the
flow of traffic by calculating the deadline miss ratio (DMR). DMR is the ratio of
those packets that do not meet their deadline at the end of invocation interval. The
operation is performed at the sink and at the end of every invocation interval the value
of DMR is communicated to the source sensor node. At the source node, the sampling
rate of packet for the next invocation interval is calculated on the basis of the previous
DMR value. This scheme does not take into account intermediate sensor nodes.
Decrease in sampling rate causes the decrease of overall throughput and thus results
in low utilization of the network. There is no provision to handle flows with different
priorities.
Different priority based scheduling schemes in [19] improve the packet
delivery ratio of important data packets. This paper presents a scenario in which
multiple sensor nodes send same type of data towards the sink. In the first scheme, the
relative importance of the packet is determined on the basis of distance of the packet
generating sensor node from the event. The nearer is the sensor node the higher the
priority of the packet generated by that sensor node and vice versa. This scheme
requires each node to determine its event distance in advance. In second scheme, the
relative importance or priority of the packet is determined on the basis of signal
strength of a sensor node to an event. The third scheme is based on timestamp and
priority is assigned on the basis of how early an event is detected by the source node.
M.Younis et. al. [20] presented an analytical overview, which highlights
123
important factors for extending the lifetime of WSN architecture. Many operational
challenges to handle the QoS traffic in sensor networks are also discussed. This paper
has reviewed many routing and MAC layer protocols that deal with the issues raised
by the constrained and unattended sensor nodes. The paper presents an energy
efficient protocol assuming data traffic with unconstrained delivery.
The mathematical expressions of queuing delay for real time and non real time
flows have been derived in [21]. The queuing delays are calculated and the results are
compared with the simulated results. The authors have implemented priority queuing
(PQ) in a sensor node. The queuing delay for two different types of traffic is
calculated by exploiting the M/G/1 queuing model. The purpose of using priority
queuing is to avoid the drawbacks associated with FIFO queuing. In this scheme a
priority queue is implemented which not only decreases the queuing delay but can
also provides preferential treatment to time sensitive applications. The authors have
not considered other QoS parameters such as packet loss, throughput, and end-to-end
delay in their work.
Figure 5.1: Classification of Health Monitoring Applications.
124
Figure 5.1 classifies health monitoring into wireless health monitoring and
context aware health monitoring applications.
5.3 Challenges and Problem Statement
Following are the main challenges while providing quality of service in
wireless sensor networks:
1. Wireless Sensor networks are normally resource constrained. Sensors are
small devices and are usually low-cost, low-power and are equipped with limited
data processing capability, transmission rate, energy, and memory. Because of the
limitations in the available bandwidth, processing power, transmission power and
the radio range of wireless channel are also limited.
2. WSNs are highly dynamic in nature. Most of the WSNs support mobility as
most of the applications in the domain of the WSNs, requires sensor nodes to be
mobile. The network topology can be changed over time due to exhausted battery
energy, node failure, node addition, node mobility, and node failure. The capacity
of channel may also change because of the dynamic adjustment of transmission
powers of the sensor/sink nodes.
3. Heterogeneity is another important challenge of WSNs. In WSNs there could
be a situation, which have different sensors node based on difference in
functionality. Because of different functionality, sensors cannot share the same
level of resource constraints. This coexistence of sensors and sinks makes WSNs
fundamentally distinct.
4. WSNs are typically operated in unpredictable environments. As it is required
to use wireless radio as the medium for data transmission when dealing with
WSNs, due to this most WSNs suffer from diverse radio interferences. Also,
query-driven and eventdriven applications can also cause the traffic load on the
network to vary unpredictably.
125
5.3.1. Problem Statement
The introduction of real time data flows and time critical data flows
specifically related to applications in medical scenarios, are required to handle data
according to time constraints associated with the flow. In the past, we have seen many
schemes in this domain which assigns priority to the data flows on the basis of the
source from which the data is coming but when we have different data flows and we
want to give priority to the packets on the basis of their criticality, then in that case we
need a mechanism which handle QoS among data flows as well as within a data flow.
In this thesis an Intelligent Application Priority Assignment Mechanism is
developed (IAPAM) which fulfils the mentioned requirements and thus helps to avoid
congestion. The mechanism increases the throughput of data critical packets as
compared to the normal packets. In order to prove its correctness and efficiency the
system is implemented and simulated using the JSIM [12] network simulator.
5.3.2. The Methodology
In the proposed scheme, an intelligent system, which decides at runtime which
data flow is to be given high priority based on its criticality, is developed. The data
generated by the source sensor nodes is compared with their respective threshold
values. If the value of any type of sensed data is greater or close to its threshold then
priority is assigned to that data stream. Different priorities can be assigned which is
based on data urgency.
At the intermediate sensor nodes a classifier is used. The task of the classifier
is to classify different data flows and assign queues according to their priority and the
source node identifier. A scheduler adaptively schedule packets on the basis of the
priority assigned by the IAPAM. Also at each intermediate node, an average queuing
delay is calculated for both priority and non-priority queues at the end of every
specific interval.
The interval is selected at which the best throughput is obtained by maintain a
threshold range TR for average queuing delay. The calculated value is compared
against the TR and an appropriate action will be taken in case the average queuing
126
delay is greater than the threshold.
5.3.3. Cross Layer Design for IAPAM
The Figure 5.2 demonstrates a cross-layer framework for wireless sensor
networks to handle real time traffic. It demonstrates a typical communication stack
and the parameters that could be shared among the layers. The design enables a higher
layer to coordinate the behaviour of lower layers. The application layer could have
parameters such as end-to-end delay and service differentiation priorities, which can
be used at routing layer and the MAC layer to provide guaranteed QoS to a specific
flow or time critical data packets.
Link Quality information that is generated at the link layer can be shared and
communicated to the MAC and the application layer. The active/idle/sleep action at
the MAC layer also helps in optimizing the performance of layers. So, it can be infer
the cross layer designs though controversial because they do not follow the classic
isolation model of layer which states that the different layers of the protocol stack
cannot share information, but still are useful to provide quality of service to hard real-
time data flows.
Figure 5.2: Cross Layer Design for IAPAM.
127
5.4 Intelligent Application Priority Assignment Mechanism
IAPAM provides prioritized scheduling to packets. In this mechanism priority
is assigned to packets of different data flows on the basis of their importance and
urgency. Scheduling is performed at the intermediate nodes on the basis of the priority
assigned to the packets and source identity. Figure 5.3 shows a sample IAPAM
scenario that contains target, sensor and sink nodes.
The target node is used to sense the data. The target nodes are moveable.
These nodes send the sensed data in the form of target packet to the sensor nodes.
TargetAgent implements the target node. This involves generating signals
(stimuli) and passing them to the lowers for transmission over the sensor channel.
TargetPacket implements the packet that represents the signal generated by a
target node.
SensorPhy receives a signal generated by the TargetAgent to get up-to-date
location of the target node and forward the generated stimulus to the sensor channel.
Sensor nodes consist of two protocol stacks. One is sensor protocol stack and
the other is wireless protocol stack. The sensor node to communicate with the target
node or to receive stimuli from the target node uses sensor protocol stack.
The sensor nodes to communicate with other intermediate sensor nodes and
sink node use the wireless protocol. There exist a middleware, which is used to
convert sensor packets into packets that are compatible with the wireless protocol
stack and can be send on the wireless channel.
128
Figure 5.3: IAPAM Network Scenario.
A sink node is developed in a plug-and-play fashion using a sensor application
layer (SensorApp), an interface layer (WirelessAgent) with the wireless protocol stack
consisting of network layer (AODV), MAC layer (MAC 802.11), and physical layer.
Some of the intermediate sensor nodes are congested because these nodes are
common to multiple target nodes. The target nodes are sending different types of data
with different priorities.
5.5 Functional Components of Application Priority Assignment
IAPAM priority assignment algorithm maintains minimum and maximum
threshold values for each type of data stream. The thresholds are predefined which are
in accordance with the values in real medical scenarios. If the data in the packet is
greater or closer to the minimum and maximum threshold limits then the packet is
considered a priority packet.
5.5.1. Packet Header Extension
The priorities are assigned at the initial level, the scheme requires a field in the
target packet so that IAPAM assign the priority to that field and this field is further
used to classify the packets. The second thing as mentioned above is that the proposed
Sensor Node
Target Node
Sink Node
129
scheme is providing priorities on the basis of data urgency and the target node from
which the data is coming (e.g. in real medical scenarios ECG data is considered more
important than temperature data). So there is need of another field in the target packet
header, which can identify the node from the data is coming. In order to fulfill the
above requirements two fields, priority and target node id, are added in the header of
target packet. The structure of the target packet used in IAPAM shown in Figure 5.4.
Figure 5.4: Structure of Target Packet.
In Figure 5.3, the target node communicates with the sensor nodes through
sensor protocol stack. Sensor nodes only accept sensor packets. At the sensor node the
target packet is encapsulated and tagged with the sensor packet header. So in order to
communicate these two fields to the sensor packet, the sensor packet header is
extended as well. In the sensor packet header the target node id field and the priority
fields are added in order to identify the source target node and the priority of the
sensor packet respectively. The structure of the sensor packet used in the proposed
scheme is presented in the Figure 5.5.
Figure 5.5: Structure of Sensor Packet.
Figure 5.6 shows that every packet is checked for three conditions. Firstly, the
packet is checked for ECG, BP or Temperature target node. Second condition checks,
130
whether the data the packet contains is greater than the minimum threshold and
thirdly, whether it is less and equal to maximum threshold of the respective flow. If
the packet fulfils all these conditions it is treated as a critical data packet otherwise as
a normal packet.
In IAPAM, 2-bit prediction scheme has been used as an intelligent mechanism
[37]. The 2-bit scheme is used to avoid branch hazards in pipelining. By using this
approach, we store the result of the last two priorities of associated packets.
The prediction must be wrong twice before it is changed i.e. if the prediction is
that the next packet will be a priority packet then the prediction is changed when two
non-priority packets arrive consecutively.
There are two thresholds for each type of data stream considered, one is
minimum threshold and the other is maximum threshold. The thresholds are
predefined which are in accordance with the values in real medical scenarios. If the
data in the packet lies between the minimum and maximum thresholds then the packet
is considered as the priority packet.
131
Figure 5.6: Priority Assignment Flowchart.
Let the minimum and maximum thresholds for the ECG, BP and Temp data
are and , and , and
respectively and the target node ids for these data flows are ,
and .
NO
NO
NO
INET PACKET
YES: RETURN ECG CRITICAL PRIORITY
YES: RETURN BP CRITICAL PRIORITY
YES: RETURN TEMP CRITICAL
PRIORITY
RETURN NORMAL PRIORITY
>TEMP MIN THS
<=TEMP MAX
THS
TEMPERATURE TARGET NODE
>BP MIN THS
<=BP MAX THS
BLOOD
PRESSURE TARGET NODE
>ECG MINTHS <=ECG MAX THS
ECG TARGET
NODE
IS ECG CRITICA
L?
IS BP CRITICA
L?
IS TEMP CRITICA
L?
IT IS NORMAL PACKET
132
The critical priority and normal priority are represented as and
respectively and the upcoming packet is represented as where ,
, and are are representing prioritized packets of
ECG, BP, Temp and normal packet from any of these nodes respectively. The
equations below represent the prioritized packets from ECG, BP, Temp and normal
packet from any of these nodes.
(5.1)
(5.2)
(5.3)
(5.4)
The flowchart diagram for the priority assignment algorithm in IAPAM is
shown in Figure 5.6.
The algorithm assigns the priority to the upcoming data packet by comparing
the data with respective thresholds. If the data falls between the threshold than critical
priority is assigned to the data packet.
133
Figure 5.7: Application Priority Assignment Functional Block.
Figure 5.7 shows flow chart of the intelligence applied in the priority
assignment mechanism. Each incoming target packet is first predicted as the priority
packet. The packet is then checked whether it falls between minimum and maximum
thresholds. In case, the packet is critical then the prediction is true. The process
continues and the predicted value remains the same. But if the prediction for the two
consecutive packets is not true then both packets are normal but the algorithm
Read next Target Packet
Predicted not Priority Packet Priority
Read next Target Packet
Predicted not Priority Packet
No
No
Yes
Yes
Read next Target Packet
Predicted Priority Packet
Read next Target Packet
Predicted Priority Packet
No
No
Yes
Yes
Priority Packet?
Priority Packet?
Priority Packet?
Priority Packet?
134
predicted these packets as critical. The value of the prediction bit is changed from
priority to non-priority.
5.5.2. Modification to Classifier
The classification of data packets is based on two fields. First is the priority of
the packet and the second is the target node id (source node of the data packet is ECG,
Blood Pressure (BP) and Temperature. The flowchart diagram of the classifier
algorithm is given in Figure 5.8.
Figure 5.8: Classifier Algorithm Design.
NO
INET PACKET
ECG TARGET NODE
CRITICAL PRIORITY
BLOOD PRESSURE TARGET NODE
CRITICAL PRIORITY
TEMPERATURE TARGET NODE
CRITICAL PRIORITY
ANY TARGET NODE
CRITICAL PRIORITY
YES: RETURN ECG QUEUE ID
NO
NO
NO
YES: RETURN BP QUEUE ID
YES: RETURN TEMP QUEUE ID
YES: RETURN NRM QUEUE ID
RETURN CONTROL QUEUE ID
CLASSIFIER
ECG PRIORITY QUEUE?
BP PRIORITY QUEUE?
TEMP PRIORITY QUEUE?
NORMAL PRIORITY QUEUE?
IT IS CONTROL PACKET
135
Each packet is first checked for the ECG target node and for its priority, if
both the conditions are satisfied then it is assigned to the highest priority queue next
to the control queue else the packet is checked for the next data flow BP and its
critical priority. The process continues and if the packet does not belong to any data
flow and is not critical, it is treated as a control packet. The flowchart describes how
the classifier selects the queue for the packets belonging to different data flows with
different priorities.
5.5.3. Interface Queue Structure
In IAPAM, after assigning priorities to packets at the node level where the
sensed data is generated, the packets are classified at each sensor node. Furthermore,
the scheme maintains different priority interface queue for each data flow as shown in
the Figure 5.9.
Figure 5.9: Virtual Queue Interface for IAPAM.
136
For example, if there are three data flows such as ECG, BP and Temperature
of patients then five interface queues are maintained. Three priority queues for the
priority packets of each data flow, one for control packets and the one for the normal
packets of every data flow. ECG data is assigned highest priority queue among the
data flows [38]. Control packets are given the highest priority among all packets of
any type. Lowest priority queue is assigned to the non-priority data. The queue
assignment to data flows is dynamic, if high priority data queue is empty then the low
priority data flows are assigned to that empty queue in order to minimize the average
queuing delay of the other data flows.
5.5.4. Queuing Delay
IAPAM calculates average queuing delay for a specific flow in a time interval
. The total queuing delay for that specific flow is calculated by adding the average
queuing delays for all time intervals. The aim is to restrict the value of average
queuing delay of number of packets below a predefined threshold value in the time
interval for each specific flow. At the end of every interval the new calculated
value of average queuing delay is compared with the threshold value. The value of
time interval is chosen on the basis of the results obtained by taking different time
intervals.
If the calculated average queuing delay of an interval is greater than the
threshold then that specific flow is assigned to higher priority queue for the next
interval so that the average queuing delay for that flow can be reduced. At the end of
this interval the average queuing delay is calculated again and is compared with the
threshold. In IAPAM, if there is an increase in the average queuing delay of ECG data
flow then the ECG data is assigned to the control queue, which has the highest
priority and the control data is assigned to the ECG queue. In the same way other data
flows can be assigned higher priority queues if their average queuing delay is higher
than the threshold.
IAPAM is using five interface queues, which can be assigned dynamically to
any data flow on the basis of the average queuing delay of each flow. The scheduler
137
first de-queue packets from the highest priority queue and if highest priority queue is
empty then the scheduler goes to the second highest priority queue and de-queue
packets. This dynamic assignment of queues among different data flows not only
reduces the average queuing delay for each specific flow but also increases overall
throughput of the network.
There are number of packets are currently present in the queue, which has
the total size . The is the time spent by each packet in the queue which is the
difference of the and , where is the time when the packet enters the queue
and is the time when it leaves it. The difference of these will give the queuing
delay for that particular packet.
(5.5)
The total queuing delay of the number of packets left the queue in a
specific time interval is given as:
(5.6)
Equation (5.6) can be written as:
(5.7)
The total queuing delay incurred by packets in an interval . By using
Equation (5.7) the average queuing delay can be computed which is given as:
at time interval (5.8)
Where gives the average queuing delay for a specific flow in a time
interval . The total queuing delay for that specific flow at a node can be calculated
138
by adding the average queuing delays of all the time intervals. Suppose, have m time
intervals then the total queuing delay for a specific flow during m time intervals is
given as:
(5.9)
The Equation (5.9) can be written as:
(5.10)
The aim is to restrict the value of average queuing delay of packets below
some predefined threshold value in the time interval for each specific flow. At the
end of every interval the new calculated value of average queuing delay is compared
with the threshold value.
The value of time interval is chosen on the basis of the results obtained by
taking different time intervals. The scheme is taking 0.2 sec as the value of the time
interval because at this value the scheme is getting the best results in terms of
throughput and delays. If the calculated average queuing delay of an interval is greater
than the threshold then the queue for that specific flow is changed to higher priority
queue for the next interval so that the average queuing delay for that flow can be
reduced. Again at the end of this interval the average queuing delay is calculated and
compared with the threshold. In IAPAM, if there is increase in the average queuing
delay of ECG data flow then the ECG data is assigned to the control queue, which has
the highest priority and the control data is assigned to the ECG queue. In the same
way the BP queue is interchanged with ECG queue, the Temp queue is interchanged
with BP queue and the queue for the normal flow is interchanged with the Temp
queue.
IAPAM is using five interface virtual queues, which can be assigned
dynamically to any data flow on the basis of the average queuing delay of each flow.
Among the five queues the queue zero is considered as the high priority queue
because the scheduler first de-queue packets from that queue and will face the
139
minimum queuing delay. If the first queue is empty then the scheduler move to the
second queue and de-queue packets. This dynamic assignment of queues among
different data flows not only reduces the average queuing delay for each specific flow
but also puts an healthy effect on the overall throughput of the network.
5.6 Performance Evaluation and Discussions
In this section, the performance of IAPAM has been evaluated. A typical
patient monitoring sensor network is simulated in JSIM [6], [12] as shown in the
Figure 5.10. Twenty sensor nodes and one sink node are distributed uniformly in a
500 × 600 square meters terrain. Six of the nodes are moving with walking speed in
random directions, while other nodes are stationary. The network consists of three
different kinds of data flows of ECG, BP and temperature generated by the sensors.
Out of twenty, six nodes are target nodes, which are generating signals or sensed data
related to the mentioned type of traffic. There are thirteen sensor nodes, which are
used to collect data from the target nodes and send it to the sink. The simulation time
is 500 sec.
Figure 5.10: Network Simulation Scenario.
140
5.6.1. Network Throughput
Figure 5.11 shows a comparison of the BWSN throughput with and without
IAPAM. The throughput of IAPAM enabled BWSN is stable and greater than the
throughput obtained in the simple scenario without IAPAM. The variance of IAPAM
enabled BWSN is 954270.1 and variance of BWSN without IAPAM is 1008683. The
mean of IAPAM enabled BWSN is 8744.5 and mean of BWSN without IAPAM is
4740.5. The small variance and mean in IAPAM enabled BWSN is because of
packets prioritization based on the urgency of the data ensuring minimal required
throughput.
Figure 5.11: Network Average Throughput Comparison.
5.6.2. Network Queuing Delay
Figure 5.12 shows average queuing delay at intermediate node near the sink.
The average queuing delay of IAPAM is varying less as compared to the simple
scenario without IAPAM. The threshold value of the average queuing delay is taken
as 0.015sec. It can be observed from simulation results that the average queuing delay
in IAPAM is below 0.015sec. So, it can be inferred that the system is performing well
in terms of queuing delay. The mean of IAPAM enabled BWSN is 0.014616 and the
mean of BWSN without IAPAM is 0.01593.
141
Figure 5.12: Network Average Queuing Delay.
The variance of BWSN without IAPAM is 3.26302E-5 and the variance of
IAPAM enabled BWSN is 7.48342E-06. The small variance and mean in IAPAM
enabled BWSN is because of priority and scheduling mechanisms of urgent data
packets.
5.6.3. Network Queuing Delay for Multiple Applications
Figure 5.13 shows the queuing delay of different types of data flows of ECG,
Blood Pressure and Temperature. The packets from the ECG nodes are incurring the
lowest average queuing delay than blood pressure and the temperature because in
hospitals ECG data is given the highest priority.
The Figure 5.14, Figure 5.15 and Figure 5.16 compare average queuing delay
for ECG, BP and Temperature respectively in BWSN with and without IAPAM. The
results clearly show that queuing delays obtained for ECG and BP are much lower
than simple scenario because ECG and BP data has higher priority as compared to
other data. Queuing Delay obtained for temperature is higher in IAPAM as compared
to simple scenario because the priority of temperature flow is low in IAPAM.
142
Figure 5.13: Comparison of Queuing Delay of Multiple Applications.
Figure 5.14: Comparison of ECG Queuing Delay.
143
Figure 5.15: Comparison of BP Queuing Delay.
Figure 5.16: Comparison of Temperature Queuing Delay.
144
5.6.4. Network Data Loss
Figure 5.17 shows the byte loss rate in IAPAM is less than the simple
scenario. This shows that the network is behaving reliably when using IAPAM. The
bytes loss ratio is greater in simple scenario as compared to IAPAM because the
priority of ECG, BP and Temp data is higher than other data and prioritize data is
scheduled with no delays and normal data is passed to normal queues resulting fewer
losses. The proposed IAPAM scheme has multiple virtual queues with different
priority levels as compared to simple scheme with one queue resulting in higher byte
loss rate. IAPAM routes the acquired data based on the urgency of the data packets.
Figure 5.17: Network Average Data Loss (Bytes).
5.6.5. Network End-to-End Delay
Figure 5.18 compares the end-to-end delay of packets for IAPAM and of
simple scenario. In IAPAM the end-to-end delay is very lower as compared to simple
scenario because of IAPAM enabled intermediate nodes.
145
Figure 5.18: Network Average End-to-End Delay.
5.6.6. Geometric Mean of the Queuing Delay
Figure 5.19 compares the geometric mean of queuing delays at different nodes
in IAPAM enabled BWSN and simple scenario. The node n1 is near to the sink and is
congested because all traffic to sink passes through it.
Figure 5.19: Geometric Mean of the Queuing Delay at Intermediate Nodes.
146
Figure 5.12 shows the average queuing delay of node n1. The average
queuing delay in IAPAM enabled BWSN at node n1 is close to the one in simple
scenario but it is still better in IAPAM. The average queuing delays at other
intermediate nodes (4, 7, 8, 9, and 12) that are away from the sink in IAPAM enabled
BWSN is less as compared to simple scenario resulting less end-to-end delay as
shown Figure 5.18.
5.6.7. ECG & TEMP Queuing Delays Comparison
Figure 5.20 presents IAPAM dynamically adjust queuing delays based on the
urgency of data. It shows that ECG has lower queuing delay & temperature has higher
queuing delays till 250 sec. After 250 sec priority of ECG data becomes normal but
the priority of temperature data increases resulting in higher delays for ECG and
lower delays for temp data.
Figure 5.20: Comparisons of ECG & TEMP Queuing Delays.
147
5.7 Conclusions
In this chapter a new data transfer scheme, called intelligent application
priority assignment mechanism (IAPAM) for biomedical wireless sensor networks
(BWSN) has been proposed, discussed, developed and simulated. IAPAM
dynamically assign priorities to individual data packets rather to a particular service or
flow by continuously monitoring queuing delays without any invoking congestion
control and avoidance mechanism. The proposed modified classifier in IAPAM
assigns data level and node level priorities to data packets.
Multiple virtual queues have been assigned to implement node level priority.
Finally, JSIM simulation model has been implemented to prove the conceptual model
of implementing the Application Controlled Handover Extension for the IAPAM in
the medical urgency scenario.
The simulation results have shown that IAPAM offers better performance than
current standards of BWSN in terms of queuing delays, end-to-end delays and
throughput. The end-to-end delay in IAPAM enabled BWSN is improved by 37.5%
as compared to the simple scenario and throughput in IAPAM enabled BWSN is 70%
higher than simple scenario. Future work will consist of implementation of a cross-
layer design that will use lower layer parameters at the routing layer so that if there is
some urgent data from different flows, it can be assigned the routes that are reliable
and congestion free.
The scheme can also be enhanced to deal with the multimedia traffic. Sending
and receiving voice and video traffic over the network and maintaining the desired
throughput and delay levels for urgent data flows is a challenging task.
5.8 References
[1] S. Koch, “Home telehealth–Current state and future trends,” Int. J. Med. Inf., Vol.
75, No. 8, pp. 565–576, 2006.
148
[2] B. Konstantas, “An overview of wearable and implantable medical sensors,” in
IMIA Yearbook, Lincoln, U.K, Vol. 2, pp. 66–69, IMIA, 2007.
[3] R. Paradiso, G. Loriga, and N. Taccini, “A wearable health care system based on
knitted integrated sensors,” IEEE Trans. Inf. Technol. Biomed., Vol. 9, No. 3, pp.
337–344, Sep. 2005.
[4] N. Maglaveras et al., “The citizen health system (CHS): A modular medical
contact center providing quality telemedicine services,” IEEE Trans. Inf. Technol.
Biomed., Vol. 9, No. 3, pp. 353–362, Sep. 2005.
[5] V. G. Koutkias, I. Chouvarda, and N. Maglaveras, “A multiagent system
enhancing home-care health services for chronic disease management,” IEEE
Trans. Inf. Technol. Biomed., Vol. 9, No. 4, pp. 528–537, Dec. 2005.
[6] N. Bricon-Souf, F. Anceaux, N. Bennani, E. Dufresne, and L. Watbled, “A
distributed coordination platform for home care: Analysis, framework and
prototype,” Int. J. Med. Inf., Vol. 74, No. 10, pp. 809–825, 2005.
[7] U. Varshney, “Pervasive healthcare and wireless health monitoring,” Mobile Netw.
Appl., Vol. 12, No. 2–3, pp. 113–127, 2007.
[8] Xeudong Liang, Ilanko Balasingham , “A QoS-aware Routing Service Framework
for Biomedical Sensor Networks”, 4th International Symposium on Wireless
Communication Systems, ISWCS07, pp. 342-345. December 2007.
[9] Xuedong Liang, Ilanko Balasingham, Performance Analysis of the IEEE802.15.4
based ECG Monitoring Network, published in IASTED Conference, 2007.
[10] Huimin She, Zhonghai Lu, Axel Jantsch, Li-Rong Zheng, Dian Zhou “A
Network-based System Architecture for Remote Medical Applications”, published
in Network Research Workshop, 2007.
[11] Ahmed Sobeih, Wei-Peng Chen, Jennifer C. Hou, Lu-Chuan Kung, Ning Li,
Hyuk Lim, Hung-Ying Tyan and Honghai Zhang “J-Sim: A Simulation and
Emulation Environment for Wireless Sensor Networks”, published in IEEE Journal,
2005.
[12] http://www.j-sim.org (December 2010)
[13] S. Patel, K. Lorincz, R. Hughes, N. Huggins, J. Growdon, D. Standaert, M. Akay,
J. Dy, M.Welsh, and P. Bonato, “Monitoring motor fluctuations in patients with
Parkinson’s disease using wearable sensors,” IEEE Transactions on Information
Technology in Biomedical, Vol. 13, No. 6, November 2009.
149
[14] M. A. Estudillo-Valderrama, L. M. Roa, J. Reina-Tosina, and D. Naranjo-
Hern´andez, “Design and implementation of a distributed fall detection system,”
IEEE Transactions on Information Technology in Biomedical, Vol. 13, No. 6,
November 2009.
[15] J. Boyle, N. Bidargaddi, A. Sarela, and M. Karunanithi, “Automatic detection of
respiration rate from ambulatory single lead ECG,” IEEE Transactions on
Information Technology in Biomedical, Vol. 13, No. 6, November 2009.
[16] J.Yoo, L.Yan, S. Lee, H. Kim, andH. J.Yoo, “A wearable ECG acquisition
system with compact planar-fashionable circuit board based shirt,”, IEEE
Transactions on Information Technology in Biomedical, Vol. 13, No. 6, November
2009.
[17] Martı´nez, J. Ferna´ ndez, M. Galarraga, L. Serrano, P. de Toledo, S. Jime´ nez-
Ferna´ ndez, S. Led, M. Martı´nez-Espronceda and J. Garcı´a, "Implementation of
an end-to-end standard-based patient monitoring solution," IET Communications,
Vol. 2, No. 2, pp. 181-191, 2008.
[18] Feng Xia, Wenhong Zhao,” Fuzzy Logic Control Based QoS Management in
Wireless Sensor /Actuator Networks”, Published in Sensors, 2007.
[19] Xianglen Yin, Hua Chen, Yang Shen , “A priority-based Packet Scheduling
Method in Wireless Sensor Networks”, published in IEEE journal, 2006.
[20] Mohamed Younis, Kemal Akkaya, Mohamed Eltoweissy, Ashraf Wadaa, “On
Handling QoS Traffic in Wireless Sensor Networks”, proceedings of the 37th
Hawaii International Conference on System Sciences - 2004.
[21] M. Y Aalsalem, Javid Taheri, Mohsin Iftikhar and Albert Y. Zomaya “Providing
QoS Guarantees to Multiple Classes of Traffic in Wireless Sensor Networks”,
published in IEEE Journal, 2008.
[22] Varshney U., "The status and future of 802.11-based WLANs," IEEE Comput,
pp. 90–93, June 2003.
[23] Calado J, Casado C, Martin A, Gomez M, Romero M, Quesada G, Lorca J, Rivas
R, "IEEE 802.11 ECG Monitoring System," In Proc. of IEEE 27th Int. EMBS
Conf, 7139–7142, Sept, 2005
[24] Yu S., Cheng J.,"A wireless physiological signal monitoring system with
integrated bluetooth and WiFi technologies," In Proc. of IEEE 27th Annual
International Conference of the EMBS, pp 2203–2206, Sept, 2005
150
[25] Lin Y., Jan I., Ko P., Chen Y., Wong J., Jan G., " A wireless PDA-based
physiological monitoring system for patient transport," IEEE Transactions on IT in
Biomedicine Vol. 8, No. 4, pp. 439–447, Dec 2005.
[26] Varshney U., "Patient monitoring using infrastructure oriented wireless LANs,"
Int. Journal on Electronic Healthcare, Vol. 2, pp. 149–163, 2006.
[27] Li F., Wu K., Reliable, "distributed and energy-efficient broadcasting in multi-
hop mobile Ad Hoc networks," In Proc. 27th IEEE Conference on Local Computer
Networks, pp. 761–769, 2002.
[28] Varshney U., "Wireless networks for patient monitoring," In Proc. Americas
Conference on Information Systems (AMCIS), August 2004.
[29] Varshney U., Sneha S.," Patient monitoring using Ad Hoc wireless networks:
reliability and power management," IEEE Communications Magazine, Vol. 44, No.
4, pp. 49-55, April 2006.
[30] Varshney U.," Addressing Un-cooperation of routers in wireless patient
monitoring," In Proc. IEEE Symposium on Computer-based Medical Systems
(CBMS), 2006.
[31] Dey A., Abowd G.," Towards a better understanding of context and context-
awareness," Tech Report, GIT-GVU-99-22, GaTech, June 1999.
[32] Chen G, Kotz D., "A survey of context-aware mobile computing research,"
Technical Report, TR2000-381, Dartmouth Computer Science, 2000.
[33] Skov M., Hoegh R., "Supporting information access in a hospital ward by a
context-aware mobile electronic patient record," Journal of Perv. and Ubiq.
Computing 10:205–214, 2006.
[34] Bardram J., "Applications of context-aware computing in hospital work—
examples and design principles," In Proc. ACM Symposium on Applied
Computing, 2004.
[35] Helal S., Mann W., Zabadani H., King J., Kaddoura Y., Jensen E.," The gator
tech smart house: a programmable pervasive space," IEEE Comput, 64–74, March
2005.
[36] Stefanov D., Bien Z., Bang W.,"The smart house for older persons and persons
with physical disabilities: structure, technology, arrangements, and perspectives,"
IEEE Trans Neural Syst. Rehabil Eng 12(2):228–250, June 2004.
151
[37] John L. Hennessy, David A. Patterson, "Computer Architecture: A Quantitative
Approach", 4th Edition, ISBN-13: 978-0123704900, Sept, 2006.
[38] Huimin She, Zhonghai Lu, Axel Jantsch, Li-Rong Zheng, Dian Zhou, “A
Network Based System Architecture for Remote Medical Applications”, Asia-
Pacific Advanced Network (APAN), China, August 2007.
152
Chapter 6
Mobility in IEEE 802.15.4 Based
IAPAM Enabled BWSN __________________________________________________________
IEEE 802.15.4 is specifically designed for low rate wireless personal area
network (LR-WPAN). It is very useful to obtain the low data rate, low power
consumption, low complexity and low cost wireless networks. It is a new technology
that has great potential for biomedical wireless sensor networks (BWSN) application
in real time such as online wireless game, mobile patient health monitoring etc. The
standard operates in two modes: one is beacon-enabled and other is non-beacon
enabled. In this chapter, (1) the performance study of non-beacon enabled modes of
IEEE 802.15.4 using IAPAM from small to large scale BWSN is evaluated, (2) the
impact of mobility on remote patient equipped with IAPAM based sensors at variable
speeds.
6.1 Introduction
Wireless sensor networks (WSNs) have gained worldwide attention in recent
years, particularly with the proliferation in Micro-Electro-Mechanical Systems
(MEMS) technology, which has facilitated the development of smart sensors. The
sensors are small with limited processing, speed, wireless communication and
resources. A wireless sensor network (WSN) is a special kind of ad hoc network that
comprises of a large number of cooperating small-scale sensors that are
geographically distributed and interconnected by wireless networks. Each sensor has
wireless communication capability and sufficient intelligence for signal processing
and networking of data. A wireless sensor network (WSN) consists of distributed
autonomous sensors to cooperatively monitor physical or environmental conditions,
such as temperature, sound, vibration, pressure, motion or pollutants. The sensor
nodes can sense, measure gather information and communicate with each other. So
153
that sensor nodes must have the abilities of sensing, computing and communicating.
Sensor nodes are expected to have lower power consumption and simple structure
characteristics. The basic mode of operation is different and also very inexpensive
from traditional networks. WSNs used in many industrial and civilian application
areas, including industrial process monitoring and control, machine health monitoring,
environment and habitat monitoring, healthcare applications, home automation, and
traffic control [1]-[3]. An example of BWSN is shown in Figure 6.1.
Figure 6.1: BWSN Proposed Architecture.
BWSN is classified into two major classes such as structured BWSN and
unstructured BWSN [1], [3]:
Structured: In a structured BWSN, all or some of the sensor nodes are deployed in a
pre-planned manner. The advantage of a structured network is that fewer nodes can be
deployed with lower network maintenance and management cost. Fewer nodes can be
deployed now since nodes are placed at specific locations to provide coverage while
ad hoc deployment can have uncovered regions.
Unstructured: An unstructured BWSN is one that contains a dense collection of
sensor nodes. Sensor nodes may be deployed in an ad hoc manner into the field. Once
deployed, the network is left unattended to perform monitoring and reporting
functions. In an unstructured BWSN, network maintenance such as managing
connectivity and detecting failures is difficult since there are so many nodes.
154
6.2 Components of BWSN
The typical architecture of the biomedical sensor node is classified in two
main components as shown in Figure 6.2.
Figure 6.2: Biomedical Sensor Node.
6.2.1. Components of Sensor Node (Hardware)
The major hardware components of a sensor node are sensors with ADC
(Analogue Digital Convertor) unit, power resource (battery), low power embedded
microprocessor, memory and radio transceiver. The microprocessor works like a
coordinator among the components of WSN. One or more sensors are used to collect
the data from environment. In BWSN the battery is used to store the temporary data.
In the context of battery there are two types of sensors such as active sensor and
passive sensor [4]:
Active Sensors: The active sensors sense the data with actively probe the
environment. They require continuous energy from a power resource.
Passive Sensors: The passive sensors sense the data without manipulating the
environment by active probing. They are self-powered to amplify their analogue
signal.
Components of Sensor Node (Software)
The major software components of a sensor node consist of five subsystems
such as Operating System Microcode, Sensor Drivers, Communication Processors,
Communication Drivers and Data Processing Mini Applications. Operating System
Sensing Unit Processing Unit
Power Unit
Mobilizer
Processor Storage Sensor ADC Transceiver
Power Generatorr
Medical Data
155
Microcode is also referred as middleware. It is a high-level software module for
supporting a variety of functions. The middleware covers the software from the
machine level functionality of the processor. Most commonly used Operating System
for WSN is Tiny OS.
Sensor Drivers are also software modules that manage the basic function for
sensor transceivers. Communication processors manage the communication functions.
The main communication functions are routing, packet buffering and forwarding,
topology maintenance, medium access control and encryption. Communication
drivers are the modules which operate the radio channel transmission link, clocking
and synchronization, signal encoding, bit recovery, bit counting, a signal levels and
modulation. Data Processing Mini Applications are the basic applications such data
processing, signal value storage etc. these application are supported at node level for
in-network processing. This layer is responsible to handle the low level operation
such as message transmission, message reception and addressing. The application
layer stays at top level. The physical (PHY) and data link layer (MAC sub-layer) is
based on IEEE 802.15.4 wireless network standard. The basic architecture is shown in
Figure 6.3.
IEEE 802.15.4 Stack Level
Application Level
Physical/Data Link Level
Figure 6.3: Network Layers of IEEE 802.15.4.
6.2.3. The Media Access Layer (MAC)
In the IEEE 802.15.4, there are two types of devices such as full-function
devices (FFD) and reduced-function devices (RFD). The FFD can talk to other RFDs
and other FFDs. FFD can operate in three modes serving as PAN coordinator,
coordinator or a device. The RFD can only talk to a FFD. In IEEE 802.15.4 standard
the LR-WPAN is consisted of a PAN coordinator and other set of devices. PAN
coordinator is the main controller of network. PAN coordinator is responsible for
initiating, maintaining and joining the network.
156
There are two main operational modes are available in the IEEE 802.15.4
network: beacon-enabled mode and non beacon- enabled mode. In the non beacon-
enabled network the beacon frames are not transmitted periodically. The beacon
request command frame is requested by device explicitly. In non beacon-enabled
mode the non-slotted CSMA/CA (Carrier Sense Multiple Access/Collision
Avoidance) is employed in the mac sublayer.
In the beacon-enabled network the beacon frames are transmitted periodically
by PAN coordinator and all communication between devices and coordinator are
based on a superframe structure. The beacon provides two services for the PAN
coordinator: PAN coordinator for synchronization with nodes uses beacon frame. The
time duration between two consecutive beacons is called Beacon Interval (BI).
In the beacon interval time, not all the time nodes can transmit data. The
beacon interval contains the active and inactive period. The active period is termed as
superframe and nodes can transmit data only during the active period (superframe).
In beacon-enabled mode the medium access is controlled by slotted
CSMA/CA (Carrier Sense Multiple Access/Collision Avoidance) technique. The
active part is divided into two periods, a contention access period (CAP) and an
optional contention free period (CFP). The optional CFP may accommodate up to
seven guaranteed time slots (GTSs) and a GTS more than one slot period. However, a
sufficient portion of CAP shall remains for contention-based access of other network
devices or new devices wishing to join the network.
157
Figure 6.4: IEEE 802.15.4 MAC Operational Modes
To access a channel during CAP a slotted CSMA/CA technique is used. All
contention based transactions must be completed before the beginning of CFP. All
transactions, which are using GTS, must be completed before the time of the next
GTS or the end of the CFP. In the superframe the network can also have GTS
(Guaranteed Tim Slot) where a time slot is dedicated to a particular node.
The CFP portion of the superframe is also call GTS. CFP can grow and shrink
depending on the number of GTSs assigned by the PAN coordinator. In GTS there is
no competition among the nodes for gaining the time slot. GTS is assigned to critical
time sensitive nodes. Figure 6.4 shows the IEEE 802.15.4 MAC operational modes.
IEEE 802.15.4 MAC
Beacon-enabled
Superframe
Contention free period (with GST)
Contention access period (without GST)
Slotted CSMA/CA
Slotted CSMA/CA
Un-Slotted CSMA/CA
Non Beacon enabled
158
6.3 Global Contribution and Related Work
In wireless sensor network the performance evaluation has been always a
challenging task. The performance evaluation techniques can be classified on
different parameters. The outdoor scenario can be different from indoor scenario. The
lowest layers of wireless sensor network use the IEEE 802.15.4 standard, but there is
different implementation of upper layers (like ZigBee).
The IEEE 802.15.4 standard [5] is most commonly used in wireless sensor
networks that exist today. The main reason of using this standard is that it uses low
power consumption, short range wireless networks, an efficient wireless connectivity
between devices, self-forming wireless communication, flexible network routing, and
unlicensed radio bands. This standard is very suitable and having attractive properties.
The IEEE 802.15.4 standard supports two layers:
• MAC(Media Access Control) sub-layer
• PHY(Physical) layer
The lower layers of the communication protocol stack are provided by IEEE
802.15.4 standard. The network administrator of wireless sensor network can define
the upper layers of the communication protocol stack. The network administrator
most commonly uses the existing standard for upper layers such as ZigBee. The
ZigBee software layer stays on the top of physical and mac sub-layer that is provided
by IEEE 802.15.4 standard. ZigBee wireless technology is most commonly used in
short-range communication system such as low-rate wireless personal area network
(LR-WPAN).
The ZigBee wireless technology is the most popular worldwide open source
standard for wireless radio sensor networks in the control and monitoring fields. In
our discussion the performance evaluation of ZigBee based wireless sensor network
are all based on the IEEE 802.15.4/ZigBee standard. There is a large number of
research papers related to the performance analysis of WSNs. Some of which will be
briefly discussed below.
159
In [6] it has been investigated that using on site PER (Packet Error Rate)
measurement, the performance of non-beacon and beacon enabled models. In
additionally, the Bluetooth interference impact over IEEE 802.15.4 is evaluated. It is
clear from results that there were losses of mode selected and Bluetooth over the PER.
Furthermore, the PER doesn’t show significance distance dependency for distances
between 3 to 10 meters.
In [7] an analytical model for the non beacon-enabled mode of the IEEE
802.15.4 medium access control (MAC) protocol is presented. A wireless sensor
network (WSN), which is, consisted of sensors nodes. These nodes transmit data to a
sink by direct links. The author reveals that by using the carrier-sense multiple access
(CSMA/CA) with collision-avoidance algorithm that is defined by IEEE 802.15.4.
The nodes transmit their packets upon reception of a query from the sink.
The evaluation of the distribution of the traffic, which is generated by the
nodes, is allowed. In particular, the probability of a node successfully accessing the
channel and the probability of a packet arriving at the sink are derived. Beside this the
evaluation of the optimum size that a packet should have allowed by the model to
maximize the success possibility for its transmission. Through simulations results are
validated. The type of channel access mechanism is allowed by the 802.15.4 standard
regarding MAC: non-beacon and beacon-enabled mode. The unslotted carrier-sense
multiple access with collision avoidance (CSMA/CA) can be used. To access the
radio channel each time a device waits for a random backoff period. After its end it
senses the channel. The device transmits the data if the channel is found to be idle, if
channel is not found to be idle then before trying to access the channel it waits for
another random period. A slotted CSMA channel-access mechanism is used by
implemented which is managed by the PAN coordinator send at regular intervals
bound the superframe.
The mobility of the sensor nodes, which is an important feature in WSN is
examined in [8]. A major challenge is to provide energy efficient and reliable
operation taking into consideration various mobility strategies. In particular due to
multi-hop nature of many WSN, local contention can reduce both the lifetime and
160
efficiency of the network and can load network and can load network-wide congestion.
Thus in [8] the interdependence between local contention and end-to-end
congestion is shown. The work is mainly based upon the slotted CSMA/CA medium
access control method, which is adopted in IEEE 802.15.4 protocol specification for
LR-WPAN. It also covers the beacon-enabled mode under various mobility scenarios
and evaluation of the beacon-enabled adopted by IEEE 802.15.4 protocol and
comprehensive performance analysis has been done. The effect of full active period
and half active period on packet delivery ratios is studied by changing the superframe
order and beacon order. The relation between congestion, mobility and local
contention is examined by changing the network traffic load. It is proved by the
results that mobility can be used to increase the performance of sensor network. The
author of the paper sys that by regulating the superframe order and beacon order to
nodes into sleep does not reduce the number of packets significantly which are
delivered to the sink in many-to-one model of communication.
In [9], several metrics, three different access schemes are evaluated and
considering the coexistence of contention free period (CFP) and contention access
period (CAP). The author is also studied the mutual influences of these two traffics. It
is shown by the results that the cost of much power consumption in terms of latency
and throughput the unslotted mode have better performance than slotted mode.
Moreover without sufficient GTS allocation the successful transmission of the CFP
frames cannot be guaranteed by the guaranteed time slot (GTS) in CFP.
In [10] the authors investigate the challenges and advantages of deploying a
single mobile sink in IEEE 802.15.4 ZigBee wireless sensor networks (WSNs). The
first part of the paper reveals the most modern research on sink mobility in WSNs. It
places a special emphasis on various types of sink mobility (controlled, random and
predictable) in WSNs. It also discusses the application scenarios, which are most
suitable for their respective development. In the second part of the paper author
presents a ZigBee-based/IEEE 802.15.4 wireless sensor network model is presented
for simulation of large scale networks. The model enables the forms of routing in the
underlying ZigBee WSN, evaluation of predictable and random sink mobility under
161
different conditions. The use of mobile sink has been recently proposed in the
literature [11]-[20]. A mobile sink consists of a wireless electronic device, which is
mounted on the top of the robot. It has the ability to move around the sensing area of
the network. The movement can be controlled or pre-programmed remotely.
Generally many advantages of deploying mobile sinks are reduced probability of sink
isolation and more balanced consumption of energy across the network (i.e. among
static nodes). The IEEE 802.15.4/ZigBee suite of standards is generally recognized as
technology of choice for applications because of their ability of cost-effective, low
power and reliable communication. The author [16]-[18] compares predictable and
random forms of sink mobility in ZigBee WSNs. The performance comparisons of
two types of ZigBee topologies and two types of sink mobility are presented in [16]-
[18]. This comparison is based on the packet loss, observed energy consumption and
packet delivery delay. The results obtained and presented suggests that generally
random sink mobility can provide better performance than predicable sink mobility,
particularly from the irrespective of average packet delay and overall energy
consumption.
In [21], a burst traffic adaptation algorithm for IEEE 802.15.4 beacon- enabled
networks is presented. The proposed algorithm is designed to accommodate the burst
traffic and to dynamically extend the active periods. The proposed algorithm is
designed for the performance enhancement in terms of energy consumption per
packet transmission and end-to-end delay. The adjustment of duty cycle is the main
aim of this algorithm in order to deal with intermittent traffic bursts. The adjustment
is made possible in an on demand manner. The device requests the extension of active
period by transmitting a busy tone, whenever it cannot transmit a packet in a given
active period at all. End-to-End delay can be shortened by the algorithm, as there is
extension of active period until the entire devices come to succeed. Due to busy tone
transmission more energy is consumed and this is also amortized by the reduction of
wastage incurred by packet collision. This work is also validated by the simulation
results in paper.
In [22], the impact of sinks mobility in WSNs is evolved. The IEEE 802.15.4
provides the mechanism to form the topology. In particular, a WSN is deployed in a
162
museum for monitoring the localization. In this work, authors analyze how a sink’s
mobility affects the connectivity and energy consumption that is used for the network
reconfiguration and formation. When a sink moves in IEEE 802.15.4, the sensors,
which are connected to it, can lose their association. The sink movement measures the
percentage of nodes connected to the network. They showed that network
connectivity can be heavily affected on it. Due to sink mobility, the loss of the
association has an impact on the network energy consumption because to restore the
association the nodes have to send energy. To avoid the performance worsening, it is
important to suitably manage the variation of connectivity and energy consumption.
The author is recommended to provide nodes with buffers suitably to prevent loss
during lack of connectivity. The IEEE 802.15.4 beacon-enabled network is sensitive
to sink mobility (mobile node) than non-beacon enabled network. When data traffic is
considered the performance of the non-beacon enabled networks must be evaluated.
In this case the mobility affects the data transfer delay and loss of packets. They
showed that when beacon-enabled networks are used, it is important to choose a
suitable value of BO (Beacon Order). This parameter affects the performance
concerning formation and reconfiguration of the network.
In [23], the authors have investigated local contention and end-to-end
congestion and their relation to mobility in WSN. A comprehensive performance
evaluation of a beacon-enabled mode adopted by IEEE 802.15.4 standard under the
different mobility scenarios. The author is studied the effect of full-active period and
half-active period on packet delivery ratios by changing the superframe order and the
beacon order. The author is also examined the relation between the local contention,
congestion and mobility by changing the network traffic load. The simulation results
prove that utilizing the mobility can increase the performance of the sensor network.
On the other side, by regulating the superframe order and beacon order to put the
nodes into sleep does not significantly reduce the number of packet delivered to sink.
In [24], a new algorithm has been proposed for beacon order adaptation in
IEEE 802.15.4 star topology network. In particular, the coordinator determines the
required duty cycle and adapts the beacon interval by observing the communication
frequency.
163
In [25], the effect of mobility is examined and the interaction between various
input parameters on the performance evaluation by using protocols designed for
wireless sensor networks. An important goal in this study is the interaction of the
MAC layer and routing protocols, which involve different mobility parameters. In
[26], the authors show that a static sink gives the best network performance. The
author has proved that if a trajectory has to be chosen for other reasons, then the
trajectory should give reasonable amount of time for each route that is a link route for
a segment of network. In case of diagonal trajectory the throughput is very low. There
are many major factors such as traffic and the type of trajectory along with node
density.
In [27], the authors have evaluated the performance of slotted CSMA/CA in
order to understand the impact of some major protocol attributes (beacon order,
superframe order and back-off exponent) on the network performance. Furthermore,
the authors have studied the application of slotted CSMA/CA for broadcast
transmission in wireless sensor networks. For large scale sensor network the backoff
algorithm of slotted CSMA/CA is not flexible. It lowers limit of the backoff delay is
always 0. This paper paves the ways to understand the slotted CSMA/CA mechanism
and its efficient used in wireless sensor networks. By introducing priority mechanism
to improve the performance of this mechanism and to turn slotted CSMA fair and
more flexible for large scale network.
In [28], the authors have investigated the performance capabilities of OPNET
Modeler [31] in simulating WSNs. The presented results show consistency with other
simulator softwares of WSNs. The author has concluded that OPNET has good
potential in simulating WSNs. Simulators can provide a vast variety of statistics and
reports at different network layer for the entire network or for an individual node.
6.4 Proposed Network Scenarios for BWSN using IAPAM
An IEEE 802.15.4 wireless sensor network (WSN) based on BWSN equipped
with IAPAM is composed of a PAN coordinator and set of sensor nodes. The
164
proposed BWSN scenario is based on PAN, which is named as Patient PAN. Each
PAN consists of one PAN coordinator and set of 50 mobile sensor nodes. In PAN the
one mobile node assumed to be a sensor attached to the patient moving at variable
speeds (5m/s, 10 m/s and 20 m/s), which is presenting a remote monitoring of a
patient at home. In the proposed network scenario the patient also moves randomly
using random waypoint mobility model. The following section briefly outlines the
scenario in detail.
6.4.1. Network Model Mobility Strategies
In our proposed IAPAM enabled BWSN the mobile node adopts the following
three main strategies:
1. Random mobility [13], [18]
2. Directed mobility [15]
3. Controlled mobility [18]
The model of BWSN can be either continuous or event based. In continuous
model the sensors periodically transmit medical data to the sink (e.g medical
monitoring application). In event-based BWSN, only a subset of sensors transmits
data to the sink at different instances of time (e.g history of blood pressure). The
selection of mobility type is very important in BWSN. The details of these mobility
strategies are discussed.
A. Random Mobility It is assumed that random walk mobility is independent of network topology
and traffic flows. In this strategy the mobile node randomly select the length and
direction of its path segment. In this type of mobility the mobile node randomly
selects the traversal time of segment and pause between the movements of two
different segments. The speed, pause time and direction are randomly selected. There
is no overhead for navigation and mobile node does not depend on the network
parameters. The randomized mobility is very effective to overcome the sink
neighbourhood problem. Due to continuous and random changes of mobile node the
energy consumption is more balanced and ultimately prolonged the lifetime of
network. From the resulting point of view the random mobility is suited for
165
continuous monitoring sensor networks. The mobile node in our proposed BWSN
used the Random WayPoint Model for random mobility.
B. Directed Mobility (predictable) In this type of mobility the mobile node always follow the same deterministic
path through the network. In this strategy, the selected path is enforced by artificial or
network obstacles in the environment. From the application point of view the
predictable mobility is suited for continuous and event based BWSN with known area
of activity. In both random and directed mobility the speed of mobile node influence
the amount of control data generated in the network and it is possibly limiting the
benefits of mobility. For example a faster mobile node moves, the more frequently it
will have to:(i) change in association of network; (ii) inform to other nodes (fixed) for
its current location, so that more overhead traffic.
The benefits of mobile node are also affected by different data rates and
periodicity of data transmission in case of random and directed mobility.
C. Controlled Mobility In this type of mobility, the path of mobile nodes becomes a function of the
current state of node’s energy consumption, network flows and to ensure optimal
network performance for all times. The speed and pause times are two other
optimization parameters for this strategy. From application point of view the
controlled mobility strategy is used in both continuous and event-based BWSN. Only
a few traffic flows are exist in event-based BWSN. The data generation is distributed
throughout the network in the continuous network. By a particular traffic flow the
location of the mobile node may not be influenced.
6.4.2. Mobility Challenges in Proposed BWSN
The use of mobile node could also impose a number of challenges towards the
network, which are:
• Due to changes in location of mobile node the routing paths need to be
updated periodically. It may causes to increase the data latency.
166
• In case of extreme mobility of mobile node may cause of increasing the packet
loss. Sometimes packet may be dropped due to outdated location.
• Due to periodic repetition of association request for reconnection may cause of
increasing the control overhead.
6.4.3. Random Mobility Model for Proposed BWSN
In this proposed work the Random Way Point (RWP) has been chosen as
mobility model. Now a detailed model is discussed in detail. A node selects its speed
from 0 to Vmax and its destination in the RWP model. The node is moving at that
speed until it reaches its destination point. After reaching at destination or
immediately starts the traveling to the next destination if the time is zero. If the speed
is low and it selects a very remote location, it takes too much time to reach at the
destination and if the speed is high, it selects a place too close to its location, it takes a
very limited time to reach the destination. This is very important point when it is
considered the length of simulation. The movement of the nodes is relevant to the
chosen location and its speed. A more suitable network is obtained for slow moving
with less pause time and the opposite is true for fast moving nodes with a relative high
pause time. If the time is equal to zero than this model is quite similar to Random
Walk (RW) mobility model. A memory less mobility pattern is represented by RW
mobility model because each step is calculation without any information of the
previous step; at regular time intervals both the speed and direction of the node are
updated such as sudden, sharp turns or sudden stops which can cause very unrealistic
behaviour.
In simulation area the mobile nodes are distributed random in the beginning of
simulation. In the center point of the simulation area the nodes have high probability
of moving in the smaller simulation areas, which can cause localized high-density
areas. Random Way Point model is very suitable for the simulation of BWSN and this
also widely used in Mobile Ad hoc Networks.
In the mobility of the mobile node an important modelling assumption that
differentiates our approach from most standard models [29]. The mobile node moves
driven by a high level mobility function, which symbolize by M. If pos is the position
167
of the mobile node in a given moment then M(pos) will return a new position pos+1
towards which the mobile node should move. This defines a trajectory for the mobile
node as a series of points pn0, pn1 = M (pn0), pn2 = M (pn1), . . . , pos = M(pos−1).
Also, the function M defines the speed sp = M (sp−1) by which the mobile node
moves from position psn −1 to position pn, the speed is bounded by a maximum value
which is depended on the scenario and models the mobility capacity of the mobile
node; this limit smax. A valid definition of M returns positions that are within the
network area and s ≤ smax. The mobility function for the mobile node can be invoked
at anytime even before reaching the designated point. The actual mechanism that
moves a mobile entity from position pn−1 to position pn it can be a human driver or
an automated navigation system.
6.4.4. GTS and CAP traffic support in BWSN (beacon enabled mode)
The proposed BWSN is also customized to include the two data traffic
generator at application layer such as Traffic Source (CAP), GTS Traffic Source
(CFP) and a Traffic Sink. The Traffic source generates acknowledged and
unacknowledged medical data during the CAP in the beacon enabled mode. The GTS
Traffic Source can produce acknowledged or unacknowledged time critical data
frames by using GTS (Guaranteed Time Slot) mechanism. The traffic sink process
module performs network statistics by receiving frames forwarded from lower layer.
The physical layer consists of IEEE 802.15.4 radio receiver (Rx) and transmitter (Tx),
the operating frequency band is 2.4 GHz and bit rate is 250 Kbps.
6.5 Simulation Environment
In order to investigate the performance analysis and impact of mobility on
remote node at variable speed in IAPAM enabled BWSN. A simulation based
extensive simulation based experimentation is conducted. Our simulation are
modelled and performed in JSIM [30] and OPNET [31]. The layout of network
scenario is captured in Figure 6.1.
In all scenarios the key simulation parameters, which are configured in
simulations, are shown in Table 6.1.
168
Table 6.1: The Simulation Parameters Simulation Parameters
Simulation Area 1000m x 1000m (square shaped)
Simulation Time 60 minutes (3600 seconds)
MAC Layer Parameters
Maximum number of retransmission 5 Acknowledge wait duration(sec) 0.05 Minimum value of the back-off exponent in the CSMA/CA
3
Maximum number of back-off the CSMA/CA algorithm will attempt before declaring a channel access failure
4
Channel sense duration (sec) 0.1 Physical Layer Parameters
Data rate(kbps) 250 Receiver sensitivity(dB) -85 Transmission band(GHz) 2.4 Transmission power(W)/Range 0.002 / 249m
Application Layer Packet interval time (sec/constant) 2 Packet size (bits/constant) 1024
6.6 Performance Evaluation and Discussions
This section compares the results of fixed nodes against mobile node in
BWSN. 10, and 50 network nodes are introduces in fixed BWSN scenario. In
comparison to that 10 and 50 mobile nodes at speed of 5 m/s, 10 m/s, 20 m/s and
random mobility are presented in second scenario. To measure the network
performance various parameters are compared:
1. Average Network Throughput
2. Average Network Delay
3. Average Network Power Consumed
6.6.1. Average Network Throughput
Figure 6.5 shows average network throughput of 10 nodes for both fixed and
mobile scenarios. It can be seen that 10 nodes (fixed) have the least network
169
throughput in comparison to 10 mobile nodes scenario. Random mobility scenario
produces highest throughput reaching an average of 80 Kbps.
Figure 6.5: Average Network Throughput (10 Nodes).
Average network throughput of 50 fixed nodes in comparison to 50 mobile
nodes with various mobility patterns is shown in Figure 6.6. 50 fixed nodes achieved
the minimum throughput of 5kbps. The highest throughput has been noticed from 50
nodes moving with 20 ms. High mobility caused the high throughput in the results. 50
nodes with 5 ms and 50 nodes with controlled mobility shown a linear behaviour in
results.
170
Figure 6.6: Average Network Throughput (50 Nodes).
6.6.2. Average Network Delay
Figure 6.7 shows average network delay of 10 nodes for both fixed and mobile
scenarios. It can be seen that 10 nodes (fixed) have the least network throughput in
comparison to 10 mobile nodes scenarios. Random mobility scenario produces
highest delay reaching an average of 0.2 seconds. Fixed nodes achieved 0.07 seconds
average end-to-end delay. Mobility scenarios with 5 ms and 20 ms showed an average
of 0.1 seconds and 0.12 seconds respectively. Random mobility is the most realistic
approach to real life mobility patterns. Deploying random mobility pattern over
proposed IAPAM enabled BWSN scenario suggests that random mobility shows
highest network delay as shown in the Figure 6.8.
171
Figure 6.7: Average Network Delay (10 Nodes).
Figure 6.8: Average Network Delay (50 Nodes).
172
6.6.3. Average Power Received
Figure 6.9 shows average network delay of mobile nodes scenarios. It has een
assumed the fixed nodes have constant power source so only scenarios based on
various mobility models has been simulated for the wireless power consumed and
received. Power is a critical issue for mobile nodes.
It can be seen that random mobility scenario received minimum power an
average of 0.06 mJouls. Mobility scenarios with 5 ms and 20 ms showed an average
of 0.1 mJouls/sec respectively.
Figure 6.9: Average Rx Power (10 Nodes).
Average network power received for 50 mobile nodes with various mobility
patterns is shown in Figure 6.10. Higher power received is directly proportional to
high volume of data sent by mobile sensor nodes. The highest power used has been
noticed from 50 nodes. High mobility caused the high throughput in the results. A
linear behaviour can be seen in results.
173
Figure 6.10: Average Rx Power (50 Nodes).
In the proposed study, a comparison of the scalability of the BWSN has been
conducted. 10 to 50 numbers of nodes has been simulated. The simulation results
show that increase the number of nodes increases the network throughput and
increases the network delay of the network. A comparison of impact of directed and
random mobility on mobile sensor node has been observed too.
6.7 Conclusions
A comparative study of different mobility patterns and their impact over
IAPAM enabled BWSN has been presented. The issue of transmitting medical data
traffic by mobile nodes has been studied and the consequences of using controlled,
directed and random mobility have been described. The simulation results have shown
that applying mobility over BWSN leads to high traffic throughput at the mobile
nodes, which causes higher average delay in the network and more power consumed.
The simulation results have shown that mobility decays the network performance. In
174
particular the simulation results presented in this chapter show that the average
network throughput for directed mobility network is 6 times more than the average
network throughput for random mobility enabled network. Furthermore, the average
network delay for directed mobility network is 2 times more then the average network
delay for random mobility enabled network
The results provide strong evidence that the variation in speed has a great impact on
the performance of the BWSN. On the basis of results obtained and presented in this
chapter it can be concluded that the random mobility of a node can be expected to
provide a better performance than directed mobility especially from the perspective of
network throughput and network delay.
6.8 References
[1] I.F. Akyildiz, W.Su, Y. Sankarasubmmaiam, E.Cayirci, “A Survey on Sensor
Networks,” IEEE Comm. Magazine, Vol. 91, No. 8, pp.102-114, Aug 2002.
[2] M.Reddy, “File:WSN.svg” in Wikipedia, Free Encyclopedia, June 2007.
http://en.wikipedia.org/wiki/File:WSN.svg, (Dec 2010).
[3] Jennifer Yick, Biswanath Mukherjee, Dipak Ghosal. Computer Networks, The
International Journal of Computer and Telecommunications Networking, Vol. 52, No.
12, August 2008.
[4] M.Reddy, “File: Sensornode.svg” in Wikipedia, Free Encyclopedia, June 2007.
http://en.wikipedia.org/wiki/File:Sensornode.svg, (Mar, 2010).
[5] IEEE Standard for Information Technology - Telecommunications and
Information Exchange Between Systems - Local and Metropolitan Area Networks –
Specific Requirements – Part 15.4: Wireless Medium Access Control (MAC) and
Physical Layer (PHY) Specifications for Low- Rate Wireless Personal Area Networks
(LR-WPANs), IEEE Std. 802.15.4-2006 (Revision of IEEE Std 802.15.4-2003),
September, 2006.
[6] Moises Manzano Herrera, Alberto Bonastre, Juan V.Capella, “Performance Study
of Non-beaconed and Beacon-Enabled Modes in IEEE 802.15.4 Under Bluetooth
Interference”, The Second International Conference on Mobile Ubiquitous Computing,
Systems, Services and Technologies, IEEE, 2008.
175
[7] Chiara Buratti, Roberto Verdone, “Performance Analysis of IEEE 802.15.4 Non
Beacon-Enabled Mode” IEEE Transactions on Vehicular Technology, Vol. 58,
September 2009.
[8] Sokullu,R.; Donertas,C.; Computer and Information Sciences, 2008. ISCIS '08.
23rd International Symposiumon, pp.1-5, 2008.
[9] ChangleLi; Huan-BangLi; Kohno,R.; Communications Workshops, 2009. ICC
Workshops 2009. IEEE International Conference, pp. 1-5, 2009.
[10] Dusan Stevanovic, Natalija Vlajic,” Performance of IEEE 802.15.4 in Wireless
Sensor Networks with a Mobile Sink Implementing Various Mobility Strategies LCN
2008, The 33rd IEEE Conference on Local Computer Networks, The Conference on
Leading Edge and Practical Computer Networking, Hyatt Regency Montreal,
Montreal, Quebec, Canada, 14-17 October 2008, Proceedings. Pages: 680-88, IEEE,
2008.
[11] S. Basagni, A. Carosi, E. Melachrinoudis, C. Petrioli and Z. M. Wang,
“Controlled sink mobility for prolonging wireless sensor networks lifetime”, Wireless
Networks, Springer Netherlands, February 2007.
[12] D. Tacconi, I. Carreras, D. Miorandi, F. Chiti, A. Casile and R. Fantacci,
“Mobile applications in disconnected WSN: System architecture, routing framework,
applications”, CREATE-NET, March 2007.
[13] M. J. Khan, W. N. Gangsterer and G. Haring, “Congestion avoidance and energy
efficient routing protocol for wireless sensor networks with a mobile sink”, Journal of
Networks, Vol. 2, No. 6, p. 42-49, December 2007.
[14] J. Luo, J. Panchard and M. Piorkowski, “MobiRoute: Routing towards a mobile
sink for improving lifetime in sensor networks”, Distributed Computing in Sensor
Systems, Vol. 40, 2006.
[15] S. Hashish and A. Karmouch, “Deployment-based solution for prolonging
network lifetime in sensor networks”, WSAN, 2008.
[16] M. Shakya, J. Zhang, P. Zhang and M. Lampe, “Design and optimization of
wireless sensor network with mobile gateway“, 21st International Conference on
Advanced Information Networking and Applications Workshops, May 2007.
[17] C. Chen and J. Ma, “Simulation study of AODV performance over IEEE
802.15.4 MAC in WSN with mobile sinks”, 21st International Conference on
Advanced Information Networking and Applications Workshops, Vol. 2, May 2007.
176
[18] Chatzigiannakis, I., Kinalis, A., and Nikoletseas, S. “Sink mobility protocols for
data collection in wireless sensor networks”, In Proceedings of the 24th Annual
Conference of the IEEE Communications Societies (INFOCOM’05), FL, USA, 2005.
[19] J. Lou and J. P Hubaux, “Joint mobility and routing for lifetime elongation in
wireless sensor networks”, In Proceedings of the 24th Annual Conference of the IEEE
Communications Societies (INFOCOM'05), FL, USA, 2005.
[20] F. Cuomo, S. D. Luna, U. Monaco and T. Melodia, “Routing in ZigBee: benefits
from exploiting the IEEE 802.15.4 association tree”, IEEE International Conference
on Communications, June 2007.
[21] Jongwook Lee; Jae Yeol Ha; Jeon,J.; Dong Sung Kim; Wook Hyun Kwon;
65th Vehicular Technology Conference, Spring, pp. 203–07, Sep, 2007.
[22] Anna Abbagnale; Emanuele Cipollone, Francesca Cuomo; ICC ’09, IEEE
International Conference on Communication 2009, Volume 53, Issue 18, pp. 57-75,
December 2009.
[23] Radosveta Sokullu, Cagdas Donertas, “Combined Effect of Mobility, Congestion
and Contention on Network Performance for IEEE 802.15.4 Based Networks”, pp.
2881-2886, May 2008.
[24] M. Neugebauer, J. Plönnigs, and K. Kabitzsch, “A new beacon order adaptation
algorithm for IEEE 802.15.4 networks,” IEEE, pp. 302-311, Aug. 2005.
[25] C. Barrett, M. Drozda, A. Marathe, and M. V. Marathe, “Characterizing the
interaction between routing and MAC protocols in ad-hoc networks,” MOBIHOC’02,
EPFL Lausanne, Switzerland, June 2002.
[26] Harsh Dhaka, Atishay Jain, Karun Verma, “Impact of Coordinator Mobility on
the throughput in a ZigBee Mesh Network”, IEEE 2nd International Advance
Computing Conference, pp. 4791-96, 2010.
[27] A. Koubaa, M. Alves, and E.Tovar, “A Comprehensive Simulation Study of
Slotted CSMA/CA for IEEE 802.15.4 Wireless Sensor Networks,” Workshop on
Factory Communication Systems (WFCS’06), Torino (Italy), Jun. 2006.
[28] I. S. Hammoodi, B. G. Stewart, A. Kocian, S. G. McMeekin, “A Comprehensive
Performance Study of OPNET Modeler For ZigBee Wireless Sensor Networks”, pp.
3786-3793, IEEE 3rd International Conference on Next Generation Mobile
Applications, Services and Technologies, 2009.
177
[29] Athanasios Kinalis, Sotiris Nikoletseas, “Scalable Data Collection Protocols for
Wireless Sensor Networks with Multiple Mobile Sinks”, 40th Annual Simulation
Symposium, 2007.
[30] http://www.j-sim.org (December 2010).
[31] www.opnet.com (December 2010).
178
Chapter 7
Conclusions and Future Work __________________________________________________________
This work has proposed an intelligent application controlled handover to
improve optical wireless converged network framework and application priority
assignment mechanism for biomedical wireless sensor networks. A new scheme for
wired wireless integrated network has been proposed in order to enhance the end-to-
end network performance. Simulation results have shown that the proposed concepts
of switching wireless interfaces, application controlled handover and interface
selection criteria have improved the performance.
7.1 Main Challenges and Solutions
The intelligent switching of radio access technology is an efficient proposed
protocol for cellular and wireless network applications. However, there are some
technical challenges required to resolve and improve this intelligent handover
mechanism. After presenting the optical wireless converged network limitations and
the existing challenges caused by applications of the current technology, this thesis
has proposed the following enhancements for the WWIN:
1. Interface Selection Process and Capabilities: In the original network
framework, the network capabilities are a set of potential features of the
device. This is not enough for a changeable interface access network such as
UWB, WiFi, WiMAX, WPAN and UMTS because these interfaces have their
set of capabilities. The type of application is a criteria required for better
interface selection process.
Proposed solution: New set of selection criteria is proposed,
developed and simulated. These criteria; such as the network coverage
and application requested, are proven to enhance the network
179
functionality in the aspects of continues function and soft handover
process as described in Chapter 3.
2. Cross Layer Application Controlled Handover: The network standards do
not consider the application type and the user-network interfaces.
Proposed solution: Introduction of the concept of cross layer to the
multiple air interfaces over fibre. This concept is based on the fact that
the APP, Transport, MAC and PHY layers communicate to enhance
network performance that is only activated by the application
controlled handover on wireless clients. In this concept, the wireless
network would be considered as a collection of high rate, low rate
wireless mobile devices surround the user at home or work.
3. Power and Application Controlled Framework.
Proposed solution: Develop the feature of distance based power
management to the network using the power control handover.
Enhanced the application controlled handover development by
introducing traditions power management scheme in the formation of
network. Wireless devices always have battery lifetime constraints. In
Chapter 4, this technology is used to introduce in parallel to intelligent
application handover to enhance network devices for longer usage and
enhancement.
4. Mesh Formation and Network Scalability and Integration.
Proposed solution: Introduce the mesh based infra for multi radio
over optical network integration features. Wireless gives you mobility,
optical fibre is more secure, and have terahertz of bandwidth. WiFi and
WiMAX have been used for mesh connectivity to router traffic through
mesh routers with a functionality of radio access units. Multihop power
controlled connectivity for requested application traffic enhances the
end-to-end network features.
180
5. Application Priority Assignment for BWSN.
Proposed solution: This work has studied the BWSN application
priority issues and the critical consequences of medical applications.
An intelligent application priority assignment mechanism has been
described, developed and proved by simulation that end-to-end delay is
reduced 37% and the throughput is increased by 70% comparing to the
normal standard of BWSN as presented in Chapter 5.
6. Mobility in IAPAM enabled BWSN.
An IEEE 802.15.4 based IAPAM enabled BWSN is composed at
variable speeds (5m/s, 10 m/s and 20 m/s) which describes a mobile
patient at hospital or home. In proposed scenarios, the mobile node
uses different mobility patterns such as random, directed and
controlled mobility models and their impact on the performance of the
BWSN.
7.2 Future Work
There are several research directions could be followed in order to enhance the
performance and functionalities of wired wireless converged network. These
directions are considered as a further step ahead for this research:
1. Convergence of 60 GHz IEEE802.15.3c and Satellite Communications:
The 60 GHz is a new WPAN technology that is required to be enhanced with the
approaches that are presented in this thesis. Power saving issue is very critical for this
technology and it should be enhanced for more network efficient and longer battery
lifetime. Routing issue should be studied in this technology and the application and
power control could be applied to accomplish further enhancement for the 60 GHz
mesh WPAN performance for applications management. Satellite communication
infrastructure could be enhanced in this work to enable the time sensitive application
for emergency and urgent communications.
2. Radio over Fiber Based Fast Handover Solutions for Vehicular
Communication System
181
The demand for intelligent transportation systems (ITSs) using the latest mobile
communication technologies continuously increase to exchange traffic information
and achieve safe, smooth, and comfortable driving. In particular, the system supports
not only voice, data but also multimedia services such as realtime video under high
mobility conditions. Since current and upcoming mobile cellular systems (e.g., GSM,
UMTS) at microwave bands cannot supply a high-speed user with such high data rate
traffic, mm-wave bands such as 36 or 60 GHz can be consideredAlthough these bands
have higher bandwidth. It leads to very small cell size (up to tens of meters) due to its
higher free space propagation loss. One promising alternative to the issue can be the
design of smart handover technique of different interfaces.
3. Security Enhancement for Optical Wireless Convergence:
More efficient security technique could be developed in order to enhance the data
encryption for more data protection. At the same time, network overload due to
heavier overhead control packets should be avoided. Network analysis with several
security technologies should lead to adapt one of the currently used security
techniques or to develop a specialised security algorithm in order to satisfy the
network security requirements according to the requirements of user applications.
Comparison study of several techniques is required using simulation model in order to
prove the proposed approach.
4. IAPAM Route Discovery and Provisioning:
A potential avenue that can extend the work in this thesis is to devise a more
functioning cross layer design that uses parameters of IAPAM at the routing layer so
that if there is some urgent data arriving from different flows then it could be assigned
those routes that are reliable and congestion free. Dynamically assignments of routes
will result in the increase of the throughput. It also assigns the prioritized data packets
those routes that are congestion free.
The scheme can also be enhanced to deal with the multimedia traffic. Sending and
receiving voice and video traffic over the network and maintaining the desired
throughput and delay levels for urgent data flows is a challenging task.
182
Publications
Published:
[1] S R Chaudhry and H S Al-Raweshidy, “Application Controlled Handover for Heterogeneous Multiple Radios over Fibre Networks”, IET Communications Vol 2, Issue 10, pp 1239-1250 November 2008.
[2] S R Chaudhry, H. S. Al-Raweshidy and S A Hussain “Impact of QAM-OFDM transmission technique over Hybrid Ricean, Rayleigh and AWGN Flat Fading Channels”, The 13th International Symposium on Wireless Personal Multimedia Communications, Towards Autonomous Network Infrastructure, (WPMC 2010), October 2010.
[3] S R Chaudhry & H S Al-Raweshidy “An Application Controlled Handover for Heterogeneous Radio in Wired Wireless Integrated Networks”, Wireless Personal Multimedia Communications IEEE (IST 07), Budapest, Hungary, July 2007.
[4] S R Chaudhry & H S Al-Raweshidy “Wireless-Optical Fibre Integration: application Service Layer Enhancement”, Wireless Personal Multimedia Communications IEEE (WPMC 06), San Diego, California, September 2006.
[5] A K Kiani, S R Chaudhry and W B Yao “An Analysis of Network Level Solutions for Mobile Multi-Homed Host”, Wireless Mobile Communications IEEE , Pakistan, April 2006.
[6] S R Chaudhry, A N Al-Khwildi, Y K Casey, H Aldelou, H S Al-Raweshidy, “WiMob Proactive and Reactive Routing Protocol Simulation Comparison”, 2nd IEEE International Conference on Information and Communication Technologies: from Theory to Applications, (ICTTA 06) Damascus, Syria, April 2006 (Invited paper)
[7] Y K Casey, A N Al-Khwildi, S R Chaudhry & H S Al-Raweshidy “Impact of Co-Interference on Bluetooth and WLAN Indoor Wireless Clusters”, Wireless Personal Multimedia Communications IEEE (WPMC 05), Denmark, September 2005.
[8] S R Chaudhry, A N Al-Khwildi, Y K Casey , H Aldelou, H S Al-Raweshidy “A System Performance Criteria of On-Demand Routing in Mobile Ad Hoc Networks” , Wireless and Mobile Computing, Networking and Communications, IEEE WiMob2005, Montreal, Canada, August 2005. (Best paper award)
[9] Talha Shahid, S R Chaudhry and H S Al-Raweshidy, “Reachability & Stability for Master Device selection in Personal Networks” WWRF14 Working Group 3 San Diego, California, USA. July 2005.
[10] S R Chaudhry, H S Al-Raweshidy and S S A Obayya ” Delay Analysis in Ad hoc On demand Distance Vector based Networks”, Multitopic Conference, 2004. Proceedings of IEEE (INMIC 2004).
[11] S R Chaudhry, H S Al-Raweshidy and S S A Obayya, “A comparative study of MB-OFDM and DS-CDMA in UWB WPAN”, WWRF November 2004.
[12] S R Chaudhry and H S Al-Raweshidy, “WIRELESS-OPTICAL FIBRE CONVERGENCE: A Gateway to Broadband Communication Networks”, Personal, Indoor and Mobile Radio Communications (PIMRC), 2006 IEEE 17th, International Symposium on September 2006.
183
[13] N. Al-Khwildi, T. Sulaiman, S R Chaudhry, Y. K. Casey and H.S. Al-Raweshdiy "A Novel On-Demand Link State Routing Protocol for Ad Hoc Wireless Networks", Fourth Generation Mobile Forum, 4GMF Annual Conference, San Diego, USA, July 2005.
Accepted:
[14] I Raza, S R Chaudhry and H S Al-Raweshidy, “Intelligent Priority Assignment Mechanism for Biomedical Wireless Sensor Networks”, IET-WSS, (Accepted Manuscript: 2011).
[15] S R Chaudhry, T. H. Sulaiman and H. S. Al-Raweshidy; “Power Controlled Scheme for Ad hoc Networks”, Springer Personal Communications. (Submitted: July 2010).
[16] S R Chaudhry, H. S. Al-Raweshidy and I Raza, “Energy Efficient Application Controlled Multi Radio PANs over Optical Network”, Vehicular Technology and Communications (Accepted in VTC May 2011).
[17] S R Chaudhry and H. S. Al-Raweshidy, “Transmitting UWB-OFDM using 16 QAM over Hybrid Flat Fading Channels”, Vehicular Technology and Communications (Accepted in VTC May 2011).
Submitted:
[18] S R Chaudhry and H S Al-Raweshidy, “Meshing Multi Radios over Fibre: A robust network architecture for Fixed Mobile Convergence”, WSEAS Transaction on Communications. (Submitted: Sep 2010)
Book Chapters:
[19] S R Chaudhry and H S Al-Raweshidy, “Application-Controlled and Power-Efficient Personal Area Network Architecture for FMC”, Fixed Mobile Convergence Handbook, ISBN No. 9781420091700, October 2010.